Are we ready for a world where AI and technology shape every corner of our lives?
In this episode of The TechEd Podcast, host Matt Kirchner sits down with Toshi Hoo, Director of the Emerging Media Lab at the Institute for the Future, to explore how technology is transforming the way we communicate, collaborate, and connect. From the breakthroughs of generative AI to the concept of the singularity, Toshi shares cutting-edge insights into what’s next for humanity—and why curiosity might be the most important skill of all.
With decades of experience in emerging technologies and strategic foresight, Toshi offers a compelling vision of a future that feels as exciting as it does uncertain. Together, Matt and Toshi unpack the promise and pitfalls of technological change, from AI’s creative potential to the ethical challenges it presents.
Listen to learn:
- A better understanding of generative AI - and why tools like ChatGPT don't actually give you "answers"
- Could the "holodeck" be more than science fiction? Toshi's work in XR and AI suggest it could be a real tool someday soon.
- Toshi's surprising connection to famed futurist Ray Kurzweil and what we know about the singularity
- How AI modeling enables more accurate scenario planning, helping organizations prepare for a range of possible futures and make smarter decisions today.
- Why curiosity isn’t just a personality trait but the defining skill for thriving in a world of rapid disrution.
3 Big Takeaways from this Episode:
- Generative AI redefines creativity but also challenges our trust in technology. Generative AI doesn’t give deterministic results, as the same inputs can yield different outputs. This non-deterministic nature enables creativity but also raises issues with reliability and accuracy. Educators should keep this in mind when having students interact with AI-driven tools in the learning experience.
- Immersive technology like XR and AI is on the verge of delivering "holodeck"-like experiences. The combination of AI and XR tools can create real-time, interactive simulations for collaboration and learning. These systems could allow users to explore environments from historical settings to molecular structures. Imagine how immersive learning can become with this technology!
- Thanks to AI, modeling and scenario planning are becoming democratized, empowering organizations to anticipate diverse futures. Modeling tools informed by AI can simulate complex systems such as city planning or healthcare data. These tools enable organizations to test strategies across multiple scenarios and adapt effectively.
Resources in this Episode:
To learn more about Institute for the Future, visit: www.iftf.org
We want to hear from you! Send us a text.
Instagram - Facebook - YouTube - TikTok - Twitter - LinkedIn
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Matt Kirchner (00:08):
Matt. It's Matt Kirchner on The TechEd Podcast,
the number one podcast in all ofSTEM at Technical Education.
Welcome back. We're gonna havean awesome episode today talking
all about the future oftechnology, as our audience
knows. I spend my fair share oftime traveling around the United
States and beyond, going toconferences, speaking at
(00:29):
conferences, sharing all thegreat things that are happening
in and around STEM and technicaleducation. And so it was that I
found myself at the Alliance forinnovation and transformation in
higher education theirconference in Phoenix, Arizona,
early August of 2024 and I mettoday's guest, and you are just
absolutely gonna love what thisindividual has to say. I
(00:50):
listened to his keynote, and Iwas just absolutely enamored
with the message we're gonnashare today with the audience of
The TechEd Podcast. Our guest isToshi, who Toshi is the director
of the Emerging Media Lab at theInstitute for the Future. And
Toshi, first of all, welcome.
It's great to see you again.
Thanks, Matt. It's great to seeyou again. So tell me about the
(01:13):
work you're doing the Institutefor the Future. I mean, that's a
pretty big name for anorganization. And of course,
you've got the title of directorof the Emerging Media Lab. But
let's just start out kind of30,000 feet Institute for the
Future. What it is that you doand why you exist? Sure.
Toshi Hoo (01:29):
Well, Institute for the Future is a non profit,
Independent Futures research,design and education
organization. We were actuallythe world's longest running
futures organization. We werefounded way back in 1968 by some
of the original computer andsocial scientists who were
working on a little project backthen called the ARPANET. We now
(01:50):
call it the internet. When itwas first started out, it was
just a government projectlooking at how you could make a
distributed communicationsystem. The vision for it was,
oh, this is just going to beused by military and academics
to have computers talk to eachother. But our founders started
saying, Well, wait a minute,what if this kind of real time
information system, what if thisscaled out? I mean, that was all
point of it is that it was couldbe so distributed that everybody
(02:12):
in the world could use it tohave, you know, real time share
of information, real timecommunication, that might change
the world, right? That mightcreate a different future than
we're imagining right now. Andof course, that's basically what
did happen. And our foundersreally have spent the next
several decades reallydeveloping a set of
methodologies around strategicforesight, as it's officially
(02:34):
called, or more informally,futures thinking. And the
institute, let's see. We worknow with large companies all
around the world, governmentagencies, large foundations,
NGOs, associations like the onewe met through a fit. We don't
actually believe that anybodycan predict the future, but we
do think it's really critical todevelop new practices, new
(02:55):
methodologies for thinking anddiscussing possible futures,
because we believe that actuallythe way the future is created is
by people thinking about it andtalking about it and making
pathways towards the futuresthey ideally the preferable
futures they'd like to see. Andthe Institute has a number of
different labs looking at allsorts of different types of
futures, everything from futureof food, future of governance,
(03:18):
we have a future of learning andeducation. The lab that I lead
and Institute for the Future iscalled the Emerging Media Lab,
as he said. And the you know,the ML, as we call it, is really
looking at the future of what Ilike to say, the future of human
communication, collaboration andconnection, all through the lens
of emerging media technology aswell as emerging media
(03:42):
mythology, meaning not just whatare the tools, but more
importantly, what are the newstories we can tell that we
couldn't tell before? What arethe new conversations we can
have that we couldn't havebefore? Also, very importantly,
who's going to be involved inthat conversation, those
conversations and thatstorytelling. So we take a very
broad definition of media to beany sort of tool or platform
(04:02):
that helps humans communicate,collaborate and connect.
Matt Kirchner (04:06):
Awesome. So you invented the internet. Is that?
What you're telling me youinvented the actually,
Toshi Hoo (04:11):
one of our founders, Paul Bannon, invented a little
thing called packet switching,which in 1968 was the idea of
wrapping data with a littlewrapper protocol that could be
handled anywhere in the world.
Coincidentally, at the sametime, same year that the
shipping container was invented.
Wow, before that, materials wereshipped around the world to kind
of random freighters, and theshipping container is exactly
(04:33):
the same idea as a data packet.
Like, let's wrap it into astandardized packet, and now it
can, like, seamlessly movearound the world. And arguably
those two inventions in 1968 thedata packet and the shipping
container, kind of led to theglobalized world of where we can
move around things andinformation. You
Matt Kirchner (04:49):
know, what a transformative time you think
about a year like 1968 I didn'trealize, by the way, that the
shipping container had beeninvented all the way back in
that period of time. You know,of course, if I'm not mistaken,
that was the same. That thatboth Bobby Kennedy and Martin
Luther King lost their lives andassassinations. And so a hugely
transformative year for theUnited States and the and the
globe. I'll tell you. There'sone other incredibly
(05:11):
transformative thing thathappened in 1968 that doesn't
get talked about as much as asit should be, and that, of
course, is the birth of MattKirk, no. And then eventually
The TechEd Podcast changing,right, exactly. So it's just an
incredibly, incrediblytransformative year. We'll give
the audience a moment just to dothe math on how old I am now
that I disclosed what year Iwas, what year I was born. But
(05:31):
what a transformative time. Andso you've been thinking about
this for a long time. At leastthe organization has. How long
have you been with theorganization? Toshi, so I joined
Institute for the Future in 2016so I just celebrated my eighth
year. Congratulations. Yeah,thank you so much. And it's been
a wonderful, weird, wonderful,amazing, unusual journey. Yeah,
(05:53):
what kind of a background leadsto a role like that as director
of the Emerging Media Lab? Well,
Toshi Hoo (05:58):
I've been interested in media and technology,
actually, from a very young age.
I got exposed to computers backin the early 70s, when I was a
little kid by my uncle, theearly computer developer and
programmer, also got exposed bythe same uncle to filmmaking,
Super Eight filmmaking when Iwas a little kid went on to just
kind of get involved infilmmaking, it, you know, and
(06:18):
animation ended up teaching alot of that as a young child,
and then eventually, when Igraduated high school, way back
in the early 90s, I got into,let's see five different film
schools, and I did them all, andnot a single one was teaching
how to use computers to makemovies. They were literally
cutting film and taping ittogether, which is cool, for
sure. But I knew at that, evenat that age, I knew that wasn't
(06:41):
the future. The next stage of mycareer was about 20 years,
actually 25 years as a mediatechnology producer. I ended up
going and teaching myselfdigital editing, because you
could at that time, I basicallywas able to get, you know, the
first version of Final Cut Proat that time where, you know, I
basically had a full multimillion dollar production studio
my backpack now we are right andlearned editing. Ended up
(07:04):
getting involved in mediaexhibits on storytelling kind of
immersive and interactive. Sogot more and more into that
whole world. And then the late90s started working for about 10
years with fame futurist RayKurzweil on a number of
different media technologyprojects, and that's where, kind
of my media technologybackground, and then kind of the
world of futures came together.
(07:27):
And I
Matt Kirchner (07:27):
followed Kurzweil for decades, actually not,
certainly not as closely as youdid. But we're going to talk
about the singularity in just alittle bit. So we'll, I hear
it's near exactly that too.
Okay, we'll get into that, andactually nearer, according
without question, we'll get intothat in just a minute and talk
about what the singular what thesingularity is and why it's
important, and why I've beenfascinated with it. And you as
well, you'll have to go back andlisten. We had Eric Newman,
(07:48):
who's the executive producer forGriselda Narcos. Oh yeah, yeah.
I'm trying to remember his otherbig Oh painkiller with Matthew
Broderick. And absolutely, it'sjust just absolutely
fascinating. He was on, I'mgonna say, maybe three months
ago or so. We'll look that up inthe show notes for the audience,
talking about the way that, youknow, he kind of came of age, I
think, right about the same timeyou did in the world of of media
(08:09):
and film and how technology isabsolutely transformed. That, in
fact, he made the predictionthat AI would write the next
great American screenplay atsome point in our lives, which I
thought was a really fascinatingthing for somebody with his
pedigree to to be talking aboutso amazing how these
technologies are changingeverything you've had a front
row seat to that all the wayback to, you know, changes in
(08:30):
film technology, working workingalongside somebody like Ray
Kurzweil, who's, you know, anabsolute international thought
leader on the technologies andthe change that we're talking
about today. What are some ofthe technologies you have your
eyes on now? I mean, what shouldour audience be thinking about
as we look to the next severalyears in terms of things that'll
really transform the way welive, work and play? Sure, well,
Toshi Hoo (08:51):
as I mentioned, I started the Emerging Media Lab
back in 2016 and prior to that,my career has been looking at
kind of different emergingtechnologies, whether that the
immersive dome shows. I mean,back in the day when just even
having interactive media was anew thing, I got very heavily
involved in that. But in 2016the big story, and kind of the
emerging technology that peoplereach in was virtual reality,
(09:12):
augmented reality. So for thefirst six years of the lab, it
was primarily focused onresearching and prototyping and
forecasting around what theimpact of xr virtual reality and
augmented reality would be on avariety of different sectors.
Having been, you know, involvedin exhibit design, I'm really
keen on focusing on immersiveexperiences. And, you know, just
(09:36):
so happens, virtual reality isan immersive experience, and
it's really difficult for peopleto understand anything about it
unless they try it right. And soa lot of what we created at
Institute for the Future overthose first six years was an
experience lab, and we curated awide range of all the kind of
emerging XR hardware, from ARmixed glasses to virtual reality
(09:58):
systems as well as. Interestingcontent. And then we also built
a lot of our own prototypes forVR and AR to try to explore
what's possible then. But yeah,the last two years, I mean, ever
since the kind of release ofchat GPT, something happened,
right? There was this new chatbot, and we'd seen chat bots
going far, as far back to 1950sAI, new idea, and there's been
(10:20):
many attempts, and there's beenthings like Siri and Alexa, but
there was a new chat bot thatcame out, chat, G, B, T, and it
was just came out as a publicresearch demo online and no
promotion. It wasn't product, noadvertising, and within a month,
it had a million users, right?
It's crazy. The reason was, isbecause it wasn't like any other
chat bot anybody had experiencedbefore, it had a level of
(10:42):
cohesiveness and generalknowledge that, by many
accounts, kind of passed theproverbial Turing test. We don't
talk about the Turing Testanymore, because we kind of
essentially passed it back then.
What was the turning test? Justso people know. So the Turing
test. Alan Turing was one of theearly innovators in computer
science, he built the famousEnigma computer, analog computer
(11:06):
for World War Two to break thecodes of the Germans. But he
imagined what all, basically allof the computing systems that we
have today. And one of histhought experiments, he imagined
this idea of a Turing test, or atest where you would have a text
interaction with a computer, andyou would have a text
interaction with the human, andyou would not be able to tell
the difference. Got it. And sofor many years, and
(11:29):
particularly, you know, when Iwas working with Ray kurzwe, he
talked a lot about, you know,when are we going to pass the
Turing test? Because it wasalways very obvious, and there
was, there is an organizationthat was running the Turing test
for many years, and different AIdevelopers were developing
systems to try to pass itessentially, but really what
we've seen now, and I thinkpeople, we've seen studies of
this, but we also can justanecdotally, I think
(11:49):
experientially, people are goingand using chat, GPT, and that's
what the amazing thing about, itproduces extremely human like
text, sure. So that was aturning point for our lab, and
we realized this. We see AI asanother medium, right? It's
going to be a medium for humancommunication, collaboration,
connection, and we've just beenfocusing on trying to digest
(12:11):
kind of the big developmentsthat have come out with this,
try to sort the noise from thehype. And a lot of my role in
Institute for the Future is asinterpreter, so helping people
understand what a technology isand its fundamentals. We've been
talking about AI for a longtime. Suddenly it's here.
Generative AI is different thanother forms of AI, and it's
(12:31):
different in some reallyimportant, non intuitive ways.
So a lot of my work is justtrying to help people kind of
wrap their mind around what thisis, especially as it changes so
quickly, right?
Matt Kirchner (12:42):
Give me an example of a non intuitive way
that AI is is different.
Toshi Hoo (12:47):
Well, probably one of the most important is that
generative AI is different fromwhat I would call classification
AI. So we've all been usingclassification AI in our daily
lives, right? That's what'ssuggesting movies to us on
Netflix, curating our socialmedia feeds and things like
that, and that's what we wouldthink of as pattern recognition.
And those algorithms are verypowerful, and they've been
(13:08):
somewhat useful, but theyhaven't been able to kind of
have cohesive conversations andgenerate new ideas. Generative
AI is much more patterngeneration. So it looks at lots
of other previous data, and whatit does is it generates new
patterns based on that data. Andthe surprising kind of discovery
(13:29):
with especially chatgpt fromOpenAI was that they were able
to not just put forth cohesiveinformation, but things that
sound kind of intelligent. Now,the non intuitive part of it is
that typically, most technologyand most software is what we
call deterministic, meaning it'sconsistent, right? You put in
certain inputs and you get thesame outputs. Generative. AI is
(13:53):
non deterministic, meaning youcan put in the same inputs and
get different outputs each time.
Imagine a calculator where youhit two plus two equals and
you've got a different answereach time, and that's what gives
it its creative ability, butit's also what makes it kind of
unpredictable at times,unreliable, right?
Hallucinations? Yeah, exactly.
(14:16):
Hallucinations is one of thebiggest issues. And so one of
the things I like to point outis that people think of they
hear AI, and they thinkprecision answers right? And the
challenge of generative AI isthat it can give correct
answers, but it's actually notdesigned to generate answers.
What generative AI is designedto do is generate output that
(14:37):
looks like an answer, and that'sconfusing, because lots of
answers that look like answersare correct answers, right.
Problem is, not only do someanswers that outputs are not
correct answers correct, youknow, factually, you know
ethically, sometimeshistorically correct, but the
problem is they look exactlylike a correct answer. Sure, and
(14:58):
that's where we start to. Getinto trouble.
Matt Kirchner (15:01):
If you look out 10 years, what will we be doing
with AI and generative AI thatwe're not doing now, that would
be like crazy.
Toshi Hoo (15:10):
Well, we're at the very, very, very, very, very
beginning, right? Where, infact, in many ways, most of us
are still using the researchdemo that came out. It wasn't
even a real product, right?
Chatgpt. We're just having alittle chat experience in a
window, right? And most of thoseare kind of linear, somewhat
ephemeral conversations, andthey're most of them are text
(15:30):
based. So what we're starting tosee already is, first of all,
the these large language modelsare not just about English,
Spanish, French or Chinese, thatlanguage is that being applied
to any way that humanscommunicate symbolically. So
that can be now we're startingto see image generators, right?
We're starting to see videogenerators, and not just
(15:50):
generators, right? These modelsare being trained by on looking
at images or video or songs or Xrays or scientific data or
molecular structures. So anyform of data is now becoming
kind of the training input intothis. So these systems are not
only able to do thoseindividually, but they can do
multi modal in combination. Soyou can have like a tech space
(16:15):
or voice based conversationabout some medical data that
it's also ingested and thenoutput a 3d animation. So that's
one of the major changes thatwe're going to see over time,
and as we kind of build out themore of the synthetic media side
of generative AI, not just thetext, but the ability to
generate not only videos andimages, but real time
(16:37):
experiences like that's when westart to approach and this is
where we start to intersect.
With something like XR, youknow, virtual and augmented
reality, right? Which is what isgoing to be the interface for
something like that, right?
Sure. And so now we're able toessentially do things that are
essentially equivalent to theholodeck, right? That's a
forecast that we had from manyyears ago that's starting to
(16:58):
become more and more feasible.
One of the ways we kind ofreframe XR and AI in the long
term is to think about this associal, spatial, generative
computing, meaning that, youknow, in the same way the
internet turned computing into asocial experience, right? We
(17:18):
just assume anything we're doingwith a computer is going to have
some connective part to it,right? It's going to be
connected to the world in realtime. So it becomes more
conversational. These tools, youknow, are going to become more
and more that. And it's going togo beyond just data, information
and media. Like right now, thechat bots we're experiencing are
giving maybe answers and advice,like it might kind of give you
(17:39):
an idea or something. The otherbig transformational aspect of
this is when, when these systemsstart to have agency, and you
might hear this term agent oragentic, and what they're
talking about is the ability todo things in the world, not just
talk about them. So that couldbe buy a plane ticket, that
could be approve alone. It couldbe, you know, operate a robot to
(18:00):
clean your kitchen. This isanother big, kind of
transformative aspect that we'restarting to see the beginnings
of as well. Now, I will saythere's a lot of visions and a
lot of hype around this idea ofcreating agents. Because the
other idea is that you could notjust credit agent to go and do
one task, but that you couldgive it kind of a high level
goal, like, can you make sure myhouse is clean and also
(18:22):
decorated for the holidays, andnot have to say, like, put up
the lights and, you know, put myclothes away and all these
things, but that it could breakdown complex series of tasks,
decide what to do, and now ithas agency and autonomy. This
vision is what is kind ofdriving a lot of this idea of
how transformative AI could bein our lives, where it's not
(18:42):
just a chat bot on a screen, butit's creating whole different
kinds of media experiences, asyou mentioned earlier, going to
transform how we create themedia that we produce, and how
we're going to create that,moving into more immersive kind
of Hall deck simulationexperiences. I mean, that's
another aspect that we talkabout quite a bit, is that. So
that's the second time you usethat term, holodeck. So I think
(19:04):
I know what you mean, but helpour audience understand what
that means. Sure, sorry, that'sa reference to a Star Trek thing
than the next generation in the90s. And the idea is that you
could step into basically, kindof a room and that create
holograms, or, you know, like avirtual reality immersive
experience, right? That it's notjust a simulation of what's
going on, but it would beinteractive. So you could say,
(19:26):
please bring me to 17th centuryChina. Okay, can recreate that
and have characters in thatspace. Or you could say, you
know, let's go inside of a COVIDmolecule and let's explore that
at scale. So if we combine kindof the idea of XR, which is the
ability to kind of go anywhereat any time, any scale, as many
times as you want, with anybody,that's kind of the simulation
(19:49):
side, right? When we look at theAI side of things, a term you
might hear a lot is model,right? Okay, with large language
models, sure, one of the ways wethink about the kind of long
term. Feature here is that we'rebuilding these capacities to
model systems. Right withlanguage models. We're modeling
kind of how language works, howintelligence works, how human
(20:11):
communication works, but we'restarting to model even more and
more complex systems, like howour bodies work, or how Alpha
fold, which is a famous projectlooked at try to anticipate how
proteins fold. How do we modelhow physical molecules interact?
And now they're working on howdrugs would interact with those.
And this is this new capacitythat, as we talk about kind of
(20:32):
transformative not just how weuse these tools, but how humans
actually even think. I call itthinking through models, right?
Traditionally, we think of likemodels, like a climate model,
right? Scientists take, youknow, really geeky scientists
have the ability to go, take allthis different kind of data, put
it into a super, super computerthat no one really has access
to, and now kind of cantheoretically anticipate what
(20:54):
the weather is going to betomorrow, or if we're going to
have climate issues. You know,in the next five years, this
capacity is about to becomedemocratized to everybody, the
ability to kind of model,meaning describe a system based
on data, so that could be a pastsystem, and then it kind of
works in hand. In hand, theability to model is the ability
(21:15):
to take information, data,observations, and make a
representation of that system.
And that system could be achemical system. It could be the
3d world that we live in. Youknow, holodeck kind of refers
the idea of like that you wouldinteract with a simulation, like
in VR characters and Lydiaenvironments, or three objects.
But now extrapolate that to muchwider forms of data, like, let's
(21:36):
say traffic data in a city, orjust urban development over
time. You could model a city,you know, and then run
simulations of either what hasalready been observed or what
could happen in the future. Andof course, that's what one of
the things we're interested inis the institute is the ability
to pre simulate scenarios from a
Matt Kirchner (21:57):
model. Awesome.
Does all this mean that we'rerapidly approaching the
singularity. What do you think?
Toshi Hoo (22:03):
Well, it depends on how you define the singularity
that Yeah. What's yourdefinition? Yeah. So the
original term, you know, Ray,kind of Ray Kurzweil, who I
worked with for a bit, he coinedthat term in more popular
vernacular. Before that, youknow, it was really an
astrophysics term that refers toan event horizon beyond which is
very difficult for humans tounderstand, because the dynamics
(22:27):
of that paradigm are sodifferent than everything else
we've ever experienced. And theclassic example of that is a
black hole, right? Because,yeah, matter, gravity, this
thing pulls light in, in waysthat are just not the way we
think about how gravity works,or how mass, what is mass?
Right? Right? When Ray kind ofintroduced it, this was kind of
(22:49):
a similar idea. And he, youknow, I think one of his biggest
contributions to kind of helpingpeople understand not just how
to think about the future, buthow transformative technology
was going to be, was this ideaof exponential change, you know,
that's a little bit more of acommon idea now, and even in
like businesses we like, weexpect, kind of like, you know,
100x you know, right? Yeah,chain transformation from
(23:12):
technology. But right? We haveto remember, back even just a
couple decades ago, change wasextremely incremental, and
technology adoption was evenreally relatively slow. Wasn't
until kind of the advent of theInternet, that you know,
software could be sharedimmediately, that that really
changed. But Ray really, kind ofput forth this idea that
technology, and this is based onkind of Moore's law, the idea
(23:34):
that, you know, the founder ofIntel said, Gordon Moore. Gordon
Moore, we double our computingpower every 18 months. And is
something he observed and wasable to kind of map out. And was
was a truism for a long time,although there's some discussion
if that we're kind of reachingour physical limits on chips,
although there might be
Matt Kirchner (23:52):
ways, well, they keep saying that, but, you know,
it's interesting, and we alwaystied it to as well, the
exponential economy and the ideathat products can double in
price performance every 12 to 18months, which I think, if I'm
not mistaken, was a kind of aKurzweil. I don't know if he
that originated with him, butthat whole idea of the
exponential economy wassomething that kind of flowed
out of the Gordon Moore ideathat we can double our ability
to process information, thatthere were the speed at which we
(24:13):
can process information every 12to 18 months, or, I think,
Moore's case, maybe every twoyears. I mean, that whole
concept just fascinates me. I'vebeen fascinated by the whole
concept of the exponentialeconomy and how quickly things
are moving, and maybe we'recoming to the the end of that
from a micro processingstandpoint, I don't know, but it
doesn't feel like we're slowingdown at all in terms of
innovation. And
Toshi Hoo (24:33):
to your point, Matt, this idea of exponential growth
has been kind of translated intoother schools of thought around
economics, around kind ofbusiness and things like that
is, of course, it's quiteappealing to the business world.
The idea of, like, exponentialgrowth of profits, or, you know,
capability, is very attractive.
So that's really kind of whatRay was talking about in his New
York Times best selling book,The singulators near that was
(24:53):
released back in the 2000s Iactually went on to co direct a
film. Film, a documentary filmbased on that book with him a
couple years later. So howclosely Did
Matt Kirchner (25:03):
you work with him? I mean, you brought him up,
and I knew that you guys havebeen connected. I followed his
work for decades, and I thinkwhat it was, I think his
original website was likeKurzweil ai.com or something. He
actually had the foresight toput the letters AI right in the
URL. And that was 20 years ago,right? Yeah, yeah.
Toshi Hoo (25:20):
So my history with Ray goes way back to the late
1970s back in Boston where Igrew up. And Ray is from the
Boston area as well. He was kindof from MIT, and my dad worked
for his company back so as ayoung child, one of Ray's kind
of biggest, earliest inventionswas the reading machine for the
blind. You know, he focused onpattern recognition at MIT, and
(25:41):
he combined optical characterrecognition system, this was in
the 1970s optical characterrecognition system with a speech
synthesis system and then somesoftware for pattern recognition
to tie it all together. Andcreated the world's first
machine that could read a bookout loud to a blind person.
Amazing. And I, as a child, mydad had one of these sales
(26:02):
records. It was, I remember, itwas one of these, like printed
records you could send out in amailer that was square. It was
kind of floppy and 45 I put iton there. I listened to the
reading machine for the blind. Iwent to his office a couple
times and got to play with itand have it read books and
stuff. So that was my first andI was actually, you asked me how
it kind of got in the futuresworld. That's actually, I
wouldn't say I was working withRay, but I was influenced by Ray
(26:22):
from a younger idea that youcould combine technologies and
that by doing that, by makingtechnology and media more
accessible, you could radicallychange the world. It wasn't till
around the late 90s, and I wasliving in the in the Boston
area, and actually my friend wasthe web developer who made
Kurzweil ai.net now it said.comIt was dot, dot, okay, yep. At
(26:44):
the time, I think he has.comhe's probably got both. Ray went
to SIGGRAPH, it's the annualcomputer graphics conference.
This was in the kind of late90s. He saw an early version of
a digital Kermit, okay, digitalpuppet. This is really early
before making digitalcharacters. Was way before Toy
(27:05):
Story and way before Pixar forsure, Henson crew had a digital
puppet and Ray saw this, and asthe mythology goes, he went home
that night had a lucid dream inwhich his female virtual Alter
Ego, Ramona, spoke to him andsaid, Ray, I want you to use
virtual reality technology tobring me to life. And so Ray
(27:26):
went to the office that day andtold some folks in the office
about this. My friend who knew Iwas doing interactive media
technologies like kind ofreached out and said, Ray has
this idea. Our other friend,Noah rafford, who went on to
become the lead futurist for didby government. By the way, I was
a 3d expert, so Noah and I werebrought in to build one of the
(27:47):
world's first real timecharacter animation systems for
Ray to perform live on the TEDstage. And wow, the year 2001 as
a female virtual author, you gohimself. So he's on stage with
this motion capture suit, andthen we translated that in real
time to a digital essentiallyavatar, a puppet. And again, no
one had heard the word avatar atCNN, right? It seems 23 years
(28:09):
ago. Yeah, crazy. We believe wasthe world's first real time
character animation performanceever done. So that was my first
kind of big project with him.
And then I went on to do a bunchof other smaller ones. But it
wasn't until around 2005 to 2008that I started working with him
on the feature film, thedocumentary,
Matt Kirchner (28:27):
got it. So is he pretty down to earth guy? Like,
I mean, you think about he'sjust so incredibly smart and so
living in the future. Tell usabout that.
Toshi Hoo (28:35):
Well, I wouldn't call Ray down to earth. He, you know,
he's a visionary. I mean, he's anice guy and he's friendly and
he's playful and He's humorous,and his family is great, and
he's actually his daughter. Ijust, actually, just saw Ray and
his daughter at their booksigning for his new book. The
Singularity Is Near. Yeah, cool.
But no, Ray was not a normalperson account. He's operating,
he's thinking kind of in adifferent sphere than most
(28:57):
folks. He's incredibly smart andhas been thinking about, kind of
like these possibilities forliterally, the last several
decades. So he isn't necessarilykind of like thinking so much
about the present. He's thinkingabout what the possibilities
are. I would kind ofcharacterize him as a bit on the
utopian side. I think he'sbelieves that, oh, more
(29:18):
technology will always be betterfor humanity. I'm not sure I
believe that. I think anybodywho he's a technology theorist,
I think anybody who is atechnology practitioner knows
that technology is always kindof broken right at number two,
and technology, the more complexyou get, the less predictable it
gets. So so living at this time,right where we're in one way
(29:38):
feeling like we're It feels likewe're living in this
technological future where wehave virtual reality and
artificial intelligence and wehave all this technological
power and potential. But it alsofeels incredibly kind of
delicate and fracture, right? No
Matt Kirchner (29:54):
question. Yeah, it was. It's interesting. I was
listening to Andrew Ng, who's,as you know, crazy entrepreneur.
Or in the crazy, successfulentrepreneur in the world of AI,
he's an adjunct professor atStanford. Watch a lot of his
stuff on YouTube, and he wastalking about the singularity,
and the interviewer asked himthe question of, are we close?
And he said, You know, I don'tknow if we'll have it in the
next 10 years. And, you know,for, I guess, for our audience
(30:16):
benefit, when I think about it,it's like we get to the point
where computers can can thinkand mimic humans in ways that
are indiscernible, which is kindof the it's not exactly the
perfect definition, but it's theway I think about it. Andrew's
answer was, I don't know ifwe'll be there in 10 years, but
I kind of hope we do. I kind ofhope we I get to see it in my
lifetime, was kind of what hesaid. And I'm not sure that
everybody looks at it exactlythe same way, you know? I think
(30:38):
it'd be cool, but I also thinkit's a little scary. Is that is
that kind of your thought?
Toshi Hoo (30:42):
I mean, the very definition of the singularity is
a paradigm that we can'tconceive of because it's going
to be so different. So ofcourse, that could be scary,
right? I mean, could bewonderful too, yeah, by
definition, right? And Ray hasthought a lot about how it could
be wonderful, but I think it'salso easy to think about how it
may not be, at very least, itmight be a different and
uncomfortable and hard to adaptto, right? I'm glad you defined
(31:03):
the singularity. I mean, there'skind of two main ideas attached
to the singularity. Number oneis what you described, that this
idea of, like, essentially humanlevel or super human level
intelligence. So the idea that,well, once we even kind of reach
human level intelligence, thefact that it's continues to
improve and self improve,potentially, right? That we read
(31:24):
this kind of exponential curve,right, and that we kind of have
take off, or lift off, as somepeople would call it, like that.
It just goes so fast beyond whatwe can think of. The other kind
of way that people think aboutthe singularity often is this
idea that sometimes referred toas transhumanism, which is that
we're merging with ourtechnology so, and that can be
(31:45):
very literal, like Elon Musk'sneural link, where we're, like,
putting electrodes in our brainand having an interface. Or it
could just be, I think it'd beargued that we're already
somewhat transhuman with ourphones, right? Like no questions
this all day. It becomes kind ofan extension of ourselves and
our body, right, in ourpsychology and our social
networks at this point. Yeah,when I get
Matt Kirchner (32:03):
that report every Sunday morning that shows my
screen time, there's no questionthat that's actually happening
to me,
Toshi Hoo (32:09):
exactly. And I think you asked about my history if,
TF, it's interesting. And when Ijoined in 2016 I also look back
to that as a time where therewas a bit of a shift, I think,
about the general public'sperception of is technology
bringing us to some sort ofutopian world. Right until that
point, it was like, oh, socialmedia, it's so fun. We're all
connected. And it was in thosefollowing years that we started
(32:29):
really seeing examples, data andreally analysis, saying, like,
maybe technology, maybe theinternet, is not great in all
ways. It's spreading massivedistant information. It's
causing overthrows of democracyin some places, right? It's a
mixed bag. And I think, youknow, that's really part of what
we try to do that city for thefuture, is to really kind of
expand our range of the types offutures. We don't when we do
(32:50):
forecasts. It's always importantnot just to consider our future,
but a range of futures. And Ithink this is a little bit of
where I diverge from rays. Rayhas a very, kind of singular
view of like, okay, we're thisis where we're going, and that's
exactly what's going to happen.
And the reality, I think, isthat the future is much more
what I call multiversal meaning,not only are there multiple
possibilities, but there's,we're all going to experience
(33:12):
that future from a multipleperspectives, right? Sure, and
those are all very different. SoI think we need to have, I mean,
I always really push for more ofkind of a pluralistic view of
like how we think about thefuture, and really try to dispel
the myth of kind of thefuturists and the ivory tower
who tell you what's going tohappen, and more that we all
need to develop our capability,individually and as collectives,
(33:34):
to the ability to kind of havecohesive and valuable
conversations about the mostimportantly, the futures that
could happen, and what are thepreferable futures we're
actually trying to create. Well,
Matt Kirchner (33:46):
and one of the things you referenced earlier
was the idea that you're notnecessarily trying to predict
the future, that nobody canpredict the future. I read a
book probably 20 years ago andsat through a series of lectures
on this idea of what they calledscenario planning. And the whole
idea was that the goal isn't tofigure out exactly what's going
to happen and get there beforeeverybody else or make sure
you're prepared for that versionof the future. Is to say, okay,
(34:09):
these are the four or fivethings that are most likely to
happen. And if I start to seethe world going in one direction
or another, and you could thinkabout that as broadly as all of
technology, or as narrowly asmaybe things that would affect
an individual business, thenI've got a plan for where to go.
And it sounds like yourphilosophy is almost the same
thing. Little
Toshi Hoo (34:28):
bit. I would take it even further to say that, you
know, I mean, we do createforecasts at the institute that
is sure, our bread and butterforecast scenarios. So forecasts
are descriptions of the future.
We like to say they're plausiblebut provocative statements about
the future that help us makebetter decisions today. Right?
It's not enough just to saysomething of the future. It's
got to be able to be somethingthat we can facilitate and use
(34:49):
today, and then scenarios aremore stories about those
futures. So how do we imaginewhat it would be like to inhabit
those as an individual? Oroperate within those future,
future forecasts as anorganization, it's through
thinking through not just theforecast. You're never going to
come to an iftf event and justread a forecast and be like,
Okay, I got I know what's goingto happen now, right? The point
(35:11):
of the actual kind of productthat I like to say is not just
even the things that couldhappen, but most importantly, as
you run through your ownscenarios around you as an
individual or you as anorganization, that you learn
about your organization, youstart to see yourself and your
organization in ways youcouldn't see before. And it's I
even caution people from thisterm future proofing, because
(35:35):
also that assumes, like, Okay, Iknow probably what's going to
happen. I think if you assumethat you know what's going to
happen, you're going to misswhat's actually happening. Yeah,
that's a really good point. Weare in a world that's
increasingly volatile andunpredictable and chaotic. It's
comforting to think, Okay, Iknow what's going to happen, or
I'm future proofed, and peoplelove that, but the comfort is
(35:57):
actually lulling us into not payattention to what's happening.
Now, in our practice, I kind ofalluded earlier, when we run a
forecast, we won what's calledthe Four alternative scenario
forecasts, which means you don'tjust run a forecast, you play
that forecast out. So forexample, if my forecast is in
five years, we are going to havemore virtual humans than real
(36:17):
humans. And then you play thatscenario out in four different
kinds of archetypes. This is aframework created by Jim data,
who is one of the grandfathersof Strategic Foresight at
University of Hawaii. And thoseare growth scenario, collapse
scenario, a constraint scenarioand a transform scenario. And
it's through running thosedifferent kinds of scenarios on
(36:39):
the assumptions that yourforecast has, that you start to
understand the dynamics of whatcould happen. So if this
happens, that might happen, andthat's going to help you better
be prepared and be what we callfuture ready, not future proof,
but increase your futurereadiness, so that you're able
to kind of re contextualize yourown model of who you are, and
(36:59):
that with the options you havein these new operational
environments, which is,
Matt Kirchner (37:04):
I mean, it's just fascinating to think about that
and to think, what were thefour, by the way, it was
transformative, collapse, growthand constraint. Yeah, so that's
an interesting way to look atthe world, without a doubt, all
the incredible work that you'redoing, at iftf, I've got to
believe you've got, like, onegreat success story of, we have
this nonprofit where there's areason that we exist that's
(37:26):
obviously beyond just generatingcash flow and generating, you
know, economic results. It'swe're doing this great work. You
have a story that might resonatein terms of where you really
help change a life or change amarket space, or do something
that you felt really good about.
I could give you a bunch ofdifferent examples
Toshi Hoo (37:44):
about kind of the work that we've done with big
companies or governmentagencies. One of the challenges
is, when you're doing 10 yearforecasts, you don't often takes
a while how to hear back frompeople if those are, you know,
helpful or useful, but we do.
Matt Kirchner (37:57):
That's the benefit too, is that if you're
wrong, nobody remembers by thetime 10 years goes by, exactly
Toshi Hoo (38:02):
I could talk about projects, and I'm happy to but
to be honest, in my work, thetimes where I feel most
successful with what I'm doingis after I give a keynote, for
example, giving a talk aboutgenerative AI. And when I come
to somebody and they tell me,thank you, you explain that to
me in a way that I could finallyunderstand. It awesome to me.
The future is less going to becreated by like, one great
(38:23):
entrepreneur that's going to bethe one that changes the world,
and more about the kind ofcollective understanding and
literacies that we're buildingacross all of our different
layers of society and differenttypes of folks. Absolutely.
Yeah, I'm happy to talk aboutproject as well, but in terms
of, like, building fitness,that's a perfect
Matt Kirchner (38:40):
example. I think it was in my life's book. And I
know you interviewed, EthanMalik at one point, where he
talks about that being one ofthe great gifts is the ability
to if and if it's not him, it'ssomebody else that I read
recently, but it talks about theability to take really complex
concepts and boil them down intosomething that the average
person, or somebody who isn't asubject matter expert in that
(39:02):
area, can understand andcomprehend and understand the
impact. And I would say youdefinitely have that gift having
seen you speak myself. So that'sa great example of how the work
that you're doing is helping tochange lives. I want to give you
one more opportunity to talkabout something that changed the
life, and that's your own life.
One of the questions we love toask every one of our guests here
on The TechEd Podcast Toshi, isto ask someone to go back in
(39:22):
time. You know, you go back intime to that 15 year old version
of yourself. You know, yourdad's working for Ray Kurzweil,
which had to be just anincredible experience to grow up
in a home like that, before youhad all the success in film,
before all you had all thesuccess in media, and then the
incredible technology stuff thatyou're doing these days, if you
could go back and give thatyoung man one piece of advice,
Toshi, what would that be?
Toshi Hoo (39:44):
That's a great question. As a futurist, I'm
often thinking about the futureand not right about the past.
You know, if I were to go backto my 15 year old self, I would
say, be your natural, curiousself. It's going to take you
places you could never imagine.
You'll go absolutely and it didyou. Know, I actually like to
say that my main product that Iproduce for the world is is
wonder, yeah, meaning motivatedcuriosity. And I think that's
(40:06):
the most transformative thingyou can do. And it's and it's
not just kind of like a playful,you know, fun feeling. It's more
helping people imagine thatmaybe the world is even larger
than they imagined. No question,it is, leads into asking
questions and growth. And Ithink that's really how the
world is really transformed. Sothat's how I got to where I'm
(40:28):
at. So if I were to say my 15year old self, I'd say, just
keep being curious. Don't try toguess where you're gonna go.
That's right,
Matt Kirchner (40:36):
never really, yeah, no, and that's so
fascinating. And I've been onthis actually, this kick about
curiosity this year, and it'sbeen influenced by guests we've
had on the podcast. So I'll giveyou just three quick examples.
Mike Bigley, who is thesuperintendent of a school
district in western Wisconsin,just created a whole emerging
technologies lab for hisstudents and all kinds of really
cool applied AI experiences forthose students when he's looking
(41:00):
for educators to be part ofleading change like that. He
said the number one personalitytrait he looks for is curiosity.
Then we had Todd wanick, who'sthe CEO of Ashley Furniture,
good friend of ours, largestfurniture manufacturer in the
world, of course. And we askedTodd, when you look for people
to lead your AI transformation?
I said, What do you look for?
And he said, Curiosity was thefirst thing that he'd look for.
(41:22):
And then we had Barbara humpton,who's the CEO of Siemens, huge
company, of course, Siemens,USA, $20 billion in revenue,
45,000 employees. Give or take.
One of the things that Barbarahumpton said is, if you have
curiosity and initiative, theworld is yours, which I thought
was really, really interestingway of looking at the world in
the fact that here you are withall the cool experiences that
(41:43):
you've had. Nobody knows whattheir career journey is going to
look like. You've had anincredible career, no doubt more
to come. But for you to say,Hey, I would remind my 15 year
old self to be curious is justone more data point on this
whole journey of curiosity thatI've been on here over the
course of 2024
Toshi Hoo (42:00):
a lot of people come up to me right now and say, What
should I tell my 10 year oldchild? Now I actually think the
future belongs to the curiouspeople. Right? We're in a period
of time where so much is beingdisrupted, and that's scary, but
it's also an incredible time fornew things to happen, and it's
the curious ones that are goingto really create and explore
those The future
Matt Kirchner (42:19):
belongs to the curious people. I can't thank
the curious Toshi, who enoughfor joining us here on The
TechEd Podcast. It's been aphenomenal conversation. We
reference some things I know ouraudience is going to want to
check out. You will find thoseand all of our show notes at
TechEd podcast.com/who that is.
TechEd podcast.com/h O, O, needto remind our audience as well
(42:42):
to check us out on social media.
As you know, we are on LinkedIn,we're on Facebook, we are on
Instagram, we are on x, we areeverywhere you would ever want
to look for your social media.
So while you're there, reach outsay hello. We would love to hear
from you, and we would love tosee you again next week on The
TechEd Podcast, thanks so muchfor being with us. You.
Matt. It's Matt Kirchner on The TechEd Podcast,
the number one podcast in all ofSTEM at Technical Education.
Welcome back. We're gonna havean awesome episode today talking
all about the future oftechnology, as our audience
knows. I spend my fair share oftime traveling around the United
States and beyond, going toconferences, speaking at
(00:29):
conferences, sharing all thegreat things that are happening
in and around STEM and technicaleducation. And so it was that I
found myself at the Alliance forinnovation and transformation in
higher education theirconference in Phoenix, Arizona,
early August of 2024 and I mettoday's guest, and you are just
absolutely gonna love what thisindividual has to say. I
(00:50):
listened to his keynote, and Iwas just absolutely enamored
with the message we're gonnashare today with the audience of
The TechEd Podcast. Our guest isToshi, who Toshi is the director
of the Emerging Media Lab at theInstitute for the Future. And
Toshi, first of all, welcome.
It's great to see you again.
Thanks, Matt. It's great to seeyou again. So tell me about the
(01:13):
work you're doing the Institutefor the Future. I mean, that's a
pretty big name for anorganization. And of course,
you've got the title of directorof the Emerging Media Lab. But
let's just start out kind of30,000 feet Institute for the
Future. What it is that you doand why you exist? Sure.
Toshi Hoo (01:29):
Well, Institute for the Future is a non profit,
Independent Futures research,design and education
organization. We were actuallythe world's longest running
futures organization. We werefounded way back in 1968 by some
of the original computer andsocial scientists who were
working on a little project backthen called the ARPANET. We now
(01:50):
call it the internet. When itwas first started out, it was
just a government projectlooking at how you could make a
distributed communicationsystem. The vision for it was,
oh, this is just going to beused by military and academics
to have computers talk to eachother. But our founders started
saying, Well, wait a minute,what if this kind of real time
information system, what if thisscaled out? I mean, that was all
point of it is that it was couldbe so distributed that everybody
(02:12):
in the world could use it tohave, you know, real time share
of information, real timecommunication, that might change
the world, right? That mightcreate a different future than
we're imagining right now. Andof course, that's basically what
did happen. And our foundersreally have spent the next
several decades reallydeveloping a set of
methodologies around strategicforesight, as it's officially
(02:34):
called, or more informally,futures thinking. And the
institute, let's see. We worknow with large companies all
around the world, governmentagencies, large foundations,
NGOs, associations like the onewe met through a fit. We don't
actually believe that anybodycan predict the future, but we
do think it's really critical todevelop new practices, new
(02:55):
methodologies for thinking anddiscussing possible futures,
because we believe that actuallythe way the future is created is
by people thinking about it andtalking about it and making
pathways towards the futuresthey ideally the preferable
futures they'd like to see. Andthe Institute has a number of
different labs looking at allsorts of different types of
futures, everything from futureof food, future of governance,
(03:18):
we have a future of learning andeducation. The lab that I lead
and Institute for the Future iscalled the Emerging Media Lab,
as he said. And the you know,the ML, as we call it, is really
looking at the future of what Ilike to say, the future of human
communication, collaboration andconnection, all through the lens
of emerging media technology aswell as emerging media
(03:42):
mythology, meaning not just whatare the tools, but more
importantly, what are the newstories we can tell that we
couldn't tell before? What arethe new conversations we can
have that we couldn't havebefore? Also, very importantly,
who's going to be involved inthat conversation, those
conversations and thatstorytelling. So we take a very
broad definition of media to beany sort of tool or platform
(04:02):
that helps humans communicate,collaborate and connect.
Matt Kirchner (04:06):
Awesome. So you invented the internet. Is that?
What you're telling me youinvented the actually,
Toshi Hoo (04:11):
one of our founders, Paul Bannon, invented a little
thing called packet switching,which in 1968 was the idea of
wrapping data with a littlewrapper protocol that could be
handled anywhere in the world.
Coincidentally, at the sametime, same year that the
shipping container was invented.
Wow, before that, materials wereshipped around the world to kind
of random freighters, and theshipping container is exactly
(04:33):
the same idea as a data packet.
Like, let's wrap it into astandardized packet, and now it
can, like, seamlessly movearound the world. And arguably
those two inventions in 1968 thedata packet and the shipping
container, kind of led to theglobalized world of where we can
move around things andinformation. You
Matt Kirchner (04:49):
know, what a transformative time you think
about a year like 1968 I didn'trealize, by the way, that the
shipping container had beeninvented all the way back in
that period of time. You know,of course, if I'm not mistaken,
that was the same. That thatboth Bobby Kennedy and Martin
Luther King lost their lives andassassinations. And so a hugely
transformative year for theUnited States and the and the
globe. I'll tell you. There'sone other incredibly
(05:11):
transformative thing thathappened in 1968 that doesn't
get talked about as much as asit should be, and that, of
course, is the birth of MattKirk, no. And then eventually
The TechEd Podcast changing,right, exactly. So it's just an
incredibly, incrediblytransformative year. We'll give
the audience a moment just to dothe math on how old I am now
that I disclosed what year Iwas, what year I was born. But
(05:31):
what a transformative time. Andso you've been thinking about
this for a long time. At leastthe organization has. How long
have you been with theorganization? Toshi, so I joined
Institute for the Future in 2016so I just celebrated my eighth
year. Congratulations. Yeah,thank you so much. And it's been
a wonderful, weird, wonderful,amazing, unusual journey. Yeah,
(05:53):
what kind of a background leadsto a role like that as director
of the Emerging Media Lab? Well,
Toshi Hoo (05:58):
I've been interested in media and technology,
actually, from a very young age.
I got exposed to computers backin the early 70s, when I was a
little kid by my uncle, theearly computer developer and
programmer, also got exposed bythe same uncle to filmmaking,
Super Eight filmmaking when Iwas a little kid went on to just
kind of get involved infilmmaking, it, you know, and
(06:18):
animation ended up teaching alot of that as a young child,
and then eventually, when Igraduated high school, way back
in the early 90s, I got into,let's see five different film
schools, and I did them all, andnot a single one was teaching
how to use computers to makemovies. They were literally
cutting film and taping ittogether, which is cool, for
sure. But I knew at that, evenat that age, I knew that wasn't
(06:41):
the future. The next stage of mycareer was about 20 years,
actually 25 years as a mediatechnology producer. I ended up
going and teaching myselfdigital editing, because you
could at that time, I basicallywas able to get, you know, the
first version of Final Cut Proat that time where, you know, I
basically had a full multimillion dollar production studio
my backpack now we are right andlearned editing. Ended up
(07:04):
getting involved in mediaexhibits on storytelling kind of
immersive and interactive. Sogot more and more into that
whole world. And then the late90s started working for about 10
years with fame futurist RayKurzweil on a number of
different media technologyprojects, and that's where, kind
of my media technologybackground, and then kind of the
world of futures came together.
(07:27):
And I
Matt Kirchner (07:27):
followed Kurzweil for decades, actually not,
certainly not as closely as youdid. But we're going to talk
about the singularity in just alittle bit. So we'll, I hear
it's near exactly that too.
Okay, we'll get into that, andactually nearer, according
without question, we'll get intothat in just a minute and talk
about what the singular what thesingularity is and why it's
important, and why I've beenfascinated with it. And you as
well, you'll have to go back andlisten. We had Eric Newman,
(07:48):
who's the executive producer forGriselda Narcos. Oh yeah, yeah.
I'm trying to remember his otherbig Oh painkiller with Matthew
Broderick. And absolutely, it'sjust just absolutely
fascinating. He was on, I'mgonna say, maybe three months
ago or so. We'll look that up inthe show notes for the audience,
talking about the way that, youknow, he kind of came of age, I
think, right about the same timeyou did in the world of of media
(08:09):
and film and how technology isabsolutely transformed. That, in
fact, he made the predictionthat AI would write the next
great American screenplay atsome point in our lives, which I
thought was a really fascinatingthing for somebody with his
pedigree to to be talking aboutso amazing how these
technologies are changingeverything you've had a front
row seat to that all the wayback to, you know, changes in
(08:30):
film technology, working workingalongside somebody like Ray
Kurzweil, who's, you know, anabsolute international thought
leader on the technologies andthe change that we're talking
about today. What are some ofthe technologies you have your
eyes on now? I mean, what shouldour audience be thinking about
as we look to the next severalyears in terms of things that'll
really transform the way welive, work and play? Sure, well,
Toshi Hoo (08:51):
as I mentioned, I started the Emerging Media Lab
back in 2016 and prior to that,my career has been looking at
kind of different emergingtechnologies, whether that the
immersive dome shows. I mean,back in the day when just even
having interactive media was anew thing, I got very heavily
involved in that. But in 2016the big story, and kind of the
emerging technology that peoplereach in was virtual reality,
(09:12):
augmented reality. So for thefirst six years of the lab, it
was primarily focused onresearching and prototyping and
forecasting around what theimpact of xr virtual reality and
augmented reality would be on avariety of different sectors.
Having been, you know, involvedin exhibit design, I'm really
keen on focusing on immersiveexperiences. And, you know, just
(09:36):
so happens, virtual reality isan immersive experience, and
it's really difficult for peopleto understand anything about it
unless they try it right. And soa lot of what we created at
Institute for the Future overthose first six years was an
experience lab, and we curated awide range of all the kind of
emerging XR hardware, from ARmixed glasses to virtual reality
(09:58):
systems as well as. Interestingcontent. And then we also built
a lot of our own prototypes forVR and AR to try to explore
what's possible then. But yeah,the last two years, I mean, ever
since the kind of release ofchat GPT, something happened,
right? There was this new chatbot, and we'd seen chat bots
going far, as far back to 1950sAI, new idea, and there's been
(10:20):
many attempts, and there's beenthings like Siri and Alexa, but
there was a new chat bot thatcame out, chat, G, B, T, and it
was just came out as a publicresearch demo online and no
promotion. It wasn't product, noadvertising, and within a month,
it had a million users, right?
It's crazy. The reason was, isbecause it wasn't like any other
chat bot anybody had experiencedbefore, it had a level of
(10:42):
cohesiveness and generalknowledge that, by many
accounts, kind of passed theproverbial Turing test. We don't
talk about the Turing Testanymore, because we kind of
essentially passed it back then.
What was the turning test? Justso people know. So the Turing
test. Alan Turing was one of theearly innovators in computer
science, he built the famousEnigma computer, analog computer
(11:06):
for World War Two to break thecodes of the Germans. But he
imagined what all, basically allof the computing systems that we
have today. And one of histhought experiments, he imagined
this idea of a Turing test, or atest where you would have a text
interaction with a computer, andyou would have a text
interaction with the human, andyou would not be able to tell
the difference. Got it. And sofor many years, and
(11:29):
particularly, you know, when Iwas working with Ray kurzwe, he
talked a lot about, you know,when are we going to pass the
Turing test? Because it wasalways very obvious, and there
was, there is an organizationthat was running the Turing test
for many years, and different AIdevelopers were developing
systems to try to pass itessentially, but really what
we've seen now, and I thinkpeople, we've seen studies of
this, but we also can justanecdotally, I think
(11:49):
experientially, people are goingand using chat, GPT, and that's
what the amazing thing about, itproduces extremely human like
text, sure. So that was aturning point for our lab, and
we realized this. We see AI asanother medium, right? It's
going to be a medium for humancommunication, collaboration,
connection, and we've just beenfocusing on trying to digest
(12:11):
kind of the big developmentsthat have come out with this,
try to sort the noise from thehype. And a lot of my role in
Institute for the Future is asinterpreter, so helping people
understand what a technology isand its fundamentals. We've been
talking about AI for a longtime. Suddenly it's here.
Generative AI is different thanother forms of AI, and it's
(12:31):
different in some reallyimportant, non intuitive ways.
So a lot of my work is justtrying to help people kind of
wrap their mind around what thisis, especially as it changes so
quickly, right?
Matt Kirchner (12:42):
Give me an example of a non intuitive way
that AI is is different.
Toshi Hoo (12:47):
Well, probably one of the most important is that
generative AI is different fromwhat I would call classification
AI. So we've all been usingclassification AI in our daily
lives, right? That's what'ssuggesting movies to us on
Netflix, curating our socialmedia feeds and things like
that, and that's what we wouldthink of as pattern recognition.
And those algorithms are verypowerful, and they've been
(13:08):
somewhat useful, but theyhaven't been able to kind of
have cohesive conversations andgenerate new ideas. Generative
AI is much more patterngeneration. So it looks at lots
of other previous data, and whatit does is it generates new
patterns based on that data. Andthe surprising kind of discovery
(13:29):
with especially chatgpt fromOpenAI was that they were able
to not just put forth cohesiveinformation, but things that
sound kind of intelligent. Now,the non intuitive part of it is
that typically, most technologyand most software is what we
call deterministic, meaning it'sconsistent, right? You put in
certain inputs and you get thesame outputs. Generative. AI is
(13:53):
non deterministic, meaning youcan put in the same inputs and
get different outputs each time.
Imagine a calculator where youhit two plus two equals and
you've got a different answereach time, and that's what gives
it its creative ability, butit's also what makes it kind of
unpredictable at times,unreliable, right?
Hallucinations? Yeah, exactly.
(14:16):
Hallucinations is one of thebiggest issues. And so one of
the things I like to point outis that people think of they
hear AI, and they thinkprecision answers right? And the
challenge of generative AI isthat it can give correct
answers, but it's actually notdesigned to generate answers.
What generative AI is designedto do is generate output that
(14:37):
looks like an answer, and that'sconfusing, because lots of
answers that look like answersare correct answers, right.
Problem is, not only do someanswers that outputs are not
correct answers correct, youknow, factually, you know
ethically, sometimeshistorically correct, but the
problem is they look exactlylike a correct answer. Sure, and
(14:58):
that's where we start to. Getinto trouble.
Matt Kirchner (15:01):
If you look out 10 years, what will we be doing
with AI and generative AI thatwe're not doing now, that would
be like crazy.
Toshi Hoo (15:10):
Well, we're at the very, very, very, very, very
beginning, right? Where, infact, in many ways, most of us
are still using the researchdemo that came out. It wasn't
even a real product, right?
Chatgpt. We're just having alittle chat experience in a
window, right? And most of thoseare kind of linear, somewhat
ephemeral conversations, andthey're most of them are text
(15:30):
based. So what we're starting tosee already is, first of all,
the these large language modelsare not just about English,
Spanish, French or Chinese, thatlanguage is that being applied
to any way that humanscommunicate symbolically. So
that can be now we're startingto see image generators, right?
We're starting to see videogenerators, and not just
(15:50):
generators, right? These modelsare being trained by on looking
at images or video or songs or Xrays or scientific data or
molecular structures. So anyform of data is now becoming
kind of the training input intothis. So these systems are not
only able to do thoseindividually, but they can do
multi modal in combination. Soyou can have like a tech space
(16:15):
or voice based conversationabout some medical data that
it's also ingested and thenoutput a 3d animation. So that's
one of the major changes thatwe're going to see over time,
and as we kind of build out themore of the synthetic media side
of generative AI, not just thetext, but the ability to
generate not only videos andimages, but real time
(16:37):
experiences like that's when westart to approach and this is
where we start to intersect.
With something like XR, youknow, virtual and augmented
reality, right? Which is what isgoing to be the interface for
something like that, right?
Sure. And so now we're able toessentially do things that are
essentially equivalent to theholodeck, right? That's a
forecast that we had from manyyears ago that's starting to
(16:58):
become more and more feasible.
One of the ways we kind ofreframe XR and AI in the long
term is to think about this associal, spatial, generative
computing, meaning that, youknow, in the same way the
internet turned computing into asocial experience, right? We
(17:18):
just assume anything we're doingwith a computer is going to have
some connective part to it,right? It's going to be
connected to the world in realtime. So it becomes more
conversational. These tools, youknow, are going to become more
and more that. And it's going togo beyond just data, information
and media. Like right now, thechat bots we're experiencing are
giving maybe answers and advice,like it might kind of give you
(17:39):
an idea or something. The otherbig transformational aspect of
this is when, when these systemsstart to have agency, and you
might hear this term agent oragentic, and what they're
talking about is the ability todo things in the world, not just
talk about them. So that couldbe buy a plane ticket, that
could be approve alone. It couldbe, you know, operate a robot to
(18:00):
clean your kitchen. This isanother big, kind of
transformative aspect that we'restarting to see the beginnings
of as well. Now, I will saythere's a lot of visions and a
lot of hype around this idea ofcreating agents. Because the
other idea is that you could notjust credit agent to go and do
one task, but that you couldgive it kind of a high level
goal, like, can you make sure myhouse is clean and also
(18:22):
decorated for the holidays, andnot have to say, like, put up
the lights and, you know, put myclothes away and all these
things, but that it could breakdown complex series of tasks,
decide what to do, and now ithas agency and autonomy. This
vision is what is kind ofdriving a lot of this idea of
how transformative AI could bein our lives, where it's not
(18:42):
just a chat bot on a screen, butit's creating whole different
kinds of media experiences, asyou mentioned earlier, going to
transform how we create themedia that we produce, and how
we're going to create that,moving into more immersive kind
of Hall deck simulationexperiences. I mean, that's
another aspect that we talkabout quite a bit, is that. So
that's the second time you usethat term, holodeck. So I think
(19:04):
I know what you mean, but helpour audience understand what
that means. Sure, sorry, that'sa reference to a Star Trek thing
than the next generation in the90s. And the idea is that you
could step into basically, kindof a room and that create
holograms, or, you know, like avirtual reality immersive
experience, right? That it's notjust a simulation of what's
going on, but it would beinteractive. So you could say,
(19:26):
please bring me to 17th centuryChina. Okay, can recreate that
and have characters in thatspace. Or you could say, you
know, let's go inside of a COVIDmolecule and let's explore that
at scale. So if we combine kindof the idea of XR, which is the
ability to kind of go anywhereat any time, any scale, as many
times as you want, with anybody,that's kind of the simulation
(19:49):
side, right? When we look at theAI side of things, a term you
might hear a lot is model,right? Okay, with large language
models, sure, one of the ways wethink about the kind of long
term. Feature here is that we'rebuilding these capacities to
model systems. Right withlanguage models. We're modeling
kind of how language works, howintelligence works, how human
(20:11):
communication works, but we'restarting to model even more and
more complex systems, like howour bodies work, or how Alpha
fold, which is a famous projectlooked at try to anticipate how
proteins fold. How do we modelhow physical molecules interact?
And now they're working on howdrugs would interact with those.
And this is this new capacitythat, as we talk about kind of
(20:32):
transformative not just how weuse these tools, but how humans
actually even think. I call itthinking through models, right?
Traditionally, we think of likemodels, like a climate model,
right? Scientists take, youknow, really geeky scientists
have the ability to go, take allthis different kind of data, put
it into a super, super computerthat no one really has access
to, and now kind of cantheoretically anticipate what
(20:54):
the weather is going to betomorrow, or if we're going to
have climate issues. You know,in the next five years, this
capacity is about to becomedemocratized to everybody, the
ability to kind of model,meaning describe a system based
on data, so that could be a pastsystem, and then it kind of
works in hand. In hand, theability to model is the ability
(21:15):
to take information, data,observations, and make a
representation of that system.
And that system could be achemical system. It could be the
3d world that we live in. Youknow, holodeck kind of refers
the idea of like that you wouldinteract with a simulation, like
in VR characters and Lydiaenvironments, or three objects.
But now extrapolate that to muchwider forms of data, like, let's
(21:36):
say traffic data in a city, orjust urban development over
time. You could model a city,you know, and then run
simulations of either what hasalready been observed or what
could happen in the future. Andof course, that's what one of
the things we're interested inis the institute is the ability
to pre simulate scenarios from a
Matt Kirchner (21:57):
model. Awesome.
Does all this mean that we'rerapidly approaching the
singularity. What do you think?
Toshi Hoo (22:03):
Well, it depends on how you define the singularity
that Yeah. What's yourdefinition? Yeah. So the
original term, you know, Ray,kind of Ray Kurzweil, who I
worked with for a bit, he coinedthat term in more popular
vernacular. Before that, youknow, it was really an
astrophysics term that refers toan event horizon beyond which is
very difficult for humans tounderstand, because the dynamics
(22:27):
of that paradigm are sodifferent than everything else
we've ever experienced. And theclassic example of that is a
black hole, right? Because,yeah, matter, gravity, this
thing pulls light in, in waysthat are just not the way we
think about how gravity works,or how mass, what is mass?
Right? Right? When Ray kind ofintroduced it, this was kind of
(22:49):
a similar idea. And he, youknow, I think one of his biggest
contributions to kind of helpingpeople understand not just how
to think about the future, buthow transformative technology
was going to be, was this ideaof exponential change, you know,
that's a little bit more of acommon idea now, and even in
like businesses we like, weexpect, kind of like, you know,
100x you know, right? Yeah,chain transformation from
(23:12):
technology. But right? We haveto remember, back even just a
couple decades ago, change wasextremely incremental, and
technology adoption was evenreally relatively slow. Wasn't
until kind of the advent of theInternet, that you know,
software could be sharedimmediately, that that really
changed. But Ray really, kind ofput forth this idea that
technology, and this is based onkind of Moore's law, the idea
(23:34):
that, you know, the founder ofIntel said, Gordon Moore. Gordon
Moore, we double our computingpower every 18 months. And is
something he observed and wasable to kind of map out. And was
was a truism for a long time,although there's some discussion
if that we're kind of reachingour physical limits on chips,
although there might be
Matt Kirchner (23:52):
ways, well, they keep saying that, but, you know,
it's interesting, and we alwaystied it to as well, the
exponential economy and the ideathat products can double in
price performance every 12 to 18months, which I think, if I'm
not mistaken, was a kind of aKurzweil. I don't know if he
that originated with him, butthat whole idea of the
exponential economy wassomething that kind of flowed
out of the Gordon Moore ideathat we can double our ability
to process information, thatthere were the speed at which we
(24:13):
can process information every 12to 18 months, or, I think,
Moore's case, maybe every twoyears. I mean, that whole
concept just fascinates me. I'vebeen fascinated by the whole
concept of the exponentialeconomy and how quickly things
are moving, and maybe we'recoming to the the end of that
from a micro processingstandpoint, I don't know, but it
doesn't feel like we're slowingdown at all in terms of
innovation. And
Toshi Hoo (24:33):
to your point, Matt, this idea of exponential growth
has been kind of translated intoother schools of thought around
economics, around kind ofbusiness and things like that
is, of course, it's quiteappealing to the business world.
The idea of, like, exponentialgrowth of profits, or, you know,
capability, is very attractive.
So that's really kind of whatRay was talking about in his New
York Times best selling book,The singulators near that was
(24:53):
released back in the 2000s Iactually went on to co direct a
film. Film, a documentary filmbased on that book with him a
couple years later. So howclosely Did
Matt Kirchner (25:03):
you work with him? I mean, you brought him up,
and I knew that you guys havebeen connected. I followed his
work for decades, and I thinkwhat it was, I think his
original website was likeKurzweil ai.com or something. He
actually had the foresight toput the letters AI right in the
URL. And that was 20 years ago,right? Yeah, yeah.
Toshi Hoo (25:20):
So my history with Ray goes way back to the late
1970s back in Boston where Igrew up. And Ray is from the
Boston area as well. He was kindof from MIT, and my dad worked
for his company back so as ayoung child, one of Ray's kind
of biggest, earliest inventionswas the reading machine for the
blind. You know, he focused onpattern recognition at MIT, and
(25:41):
he combined optical characterrecognition system, this was in
the 1970s optical characterrecognition system with a speech
synthesis system and then somesoftware for pattern recognition
to tie it all together. Andcreated the world's first
machine that could read a bookout loud to a blind person.
Amazing. And I, as a child, mydad had one of these sales
(26:02):
records. It was, I remember, itwas one of these, like printed
records you could send out in amailer that was square. It was
kind of floppy and 45 I put iton there. I listened to the
reading machine for the blind. Iwent to his office a couple
times and got to play with itand have it read books and
stuff. So that was my first andI was actually, you asked me how
it kind of got in the futuresworld. That's actually, I
wouldn't say I was working withRay, but I was influenced by Ray
(26:22):
from a younger idea that youcould combine technologies and
that by doing that, by makingtechnology and media more
accessible, you could radicallychange the world. It wasn't till
around the late 90s, and I wasliving in the in the Boston
area, and actually my friend wasthe web developer who made
Kurzweil ai.net now it said.comIt was dot, dot, okay, yep. At
(26:44):
the time, I think he has.comhe's probably got both. Ray went
to SIGGRAPH, it's the annualcomputer graphics conference.
This was in the kind of late90s. He saw an early version of
a digital Kermit, okay, digitalpuppet. This is really early
before making digitalcharacters. Was way before Toy
(27:05):
Story and way before Pixar forsure, Henson crew had a digital
puppet and Ray saw this, and asthe mythology goes, he went home
that night had a lucid dream inwhich his female virtual Alter
Ego, Ramona, spoke to him andsaid, Ray, I want you to use
virtual reality technology tobring me to life. And so Ray
(27:26):
went to the office that day andtold some folks in the office
about this. My friend who knew Iwas doing interactive media
technologies like kind ofreached out and said, Ray has
this idea. Our other friend,Noah rafford, who went on to
become the lead futurist for didby government. By the way, I was
a 3d expert, so Noah and I werebrought in to build one of the
(27:47):
world's first real timecharacter animation systems for
Ray to perform live on the TEDstage. And wow, the year 2001 as
a female virtual author, you gohimself. So he's on stage with
this motion capture suit, andthen we translated that in real
time to a digital essentiallyavatar, a puppet. And again, no
one had heard the word avatar atCNN, right? It seems 23 years
(28:09):
ago. Yeah, crazy. We believe wasthe world's first real time
character animation performanceever done. So that was my first
kind of big project with him.
And then I went on to do a bunchof other smaller ones. But it
wasn't until around 2005 to 2008that I started working with him
on the feature film, thedocumentary,
Matt Kirchner (28:27):
got it. So is he pretty down to earth guy? Like,
I mean, you think about he'sjust so incredibly smart and so
living in the future. Tell usabout that.
Toshi Hoo (28:35):
Well, I wouldn't call Ray down to earth. He, you know,
he's a visionary. I mean, he's anice guy and he's friendly and
he's playful and He's humorous,and his family is great, and
he's actually his daughter. Ijust, actually, just saw Ray and
his daughter at their booksigning for his new book. The
Singularity Is Near. Yeah, cool.
But no, Ray was not a normalperson account. He's operating,
he's thinking kind of in adifferent sphere than most
(28:57):
folks. He's incredibly smart andhas been thinking about, kind of
like these possibilities forliterally, the last several
decades. So he isn't necessarilykind of like thinking so much
about the present. He's thinkingabout what the possibilities
are. I would kind ofcharacterize him as a bit on the
utopian side. I think he'sbelieves that, oh, more
(29:18):
technology will always be betterfor humanity. I'm not sure I
believe that. I think anybodywho he's a technology theorist,
I think anybody who is atechnology practitioner knows
that technology is always kindof broken right at number two,
and technology, the more complexyou get, the less predictable it
gets. So so living at this time,right where we're in one way
(29:38):
feeling like we're It feels likewe're living in this
technological future where wehave virtual reality and
artificial intelligence and wehave all this technological
power and potential. But it alsofeels incredibly kind of
delicate and fracture, right? No
Matt Kirchner (29:54):
question. Yeah, it was. It's interesting. I was
listening to Andrew Ng, who's,as you know, crazy entrepreneur.
Or in the crazy, successfulentrepreneur in the world of AI,
he's an adjunct professor atStanford. Watch a lot of his
stuff on YouTube, and he wastalking about the singularity,
and the interviewer asked himthe question of, are we close?
And he said, You know, I don'tknow if we'll have it in the
next 10 years. And, you know,for, I guess, for our audience
(30:16):
benefit, when I think about it,it's like we get to the point
where computers can can thinkand mimic humans in ways that
are indiscernible, which is kindof the it's not exactly the
perfect definition, but it's theway I think about it. Andrew's
answer was, I don't know ifwe'll be there in 10 years, but
I kind of hope we do. I kind ofhope we I get to see it in my
lifetime, was kind of what hesaid. And I'm not sure that
everybody looks at it exactlythe same way, you know? I think
(30:38):
it'd be cool, but I also thinkit's a little scary. Is that is
that kind of your thought?
Toshi Hoo (30:42):
I mean, the very definition of the singularity is
a paradigm that we can'tconceive of because it's going
to be so different. So ofcourse, that could be scary,
right? I mean, could bewonderful too, yeah, by
definition, right? And Ray hasthought a lot about how it could
be wonderful, but I think it'salso easy to think about how it
may not be, at very least, itmight be a different and
uncomfortable and hard to adaptto, right? I'm glad you defined
(31:03):
the singularity. I mean, there'skind of two main ideas attached
to the singularity. Number oneis what you described, that this
idea of, like, essentially humanlevel or super human level
intelligence. So the idea that,well, once we even kind of reach
human level intelligence, thefact that it's continues to
improve and self improve,potentially, right? That we read
(31:24):
this kind of exponential curve,right, and that we kind of have
take off, or lift off, as somepeople would call it, like that.
It just goes so fast beyond whatwe can think of. The other kind
of way that people think aboutthe singularity often is this
idea that sometimes referred toas transhumanism, which is that
we're merging with ourtechnology so, and that can be
(31:45):
very literal, like Elon Musk'sneural link, where we're, like,
putting electrodes in our brainand having an interface. Or it
could just be, I think it'd beargued that we're already
somewhat transhuman with ourphones, right? Like no questions
this all day. It becomes kind ofan extension of ourselves and
our body, right, in ourpsychology and our social
networks at this point. Yeah,when I get
Matt Kirchner (32:03):
that report every Sunday morning that shows my
screen time, there's no questionthat that's actually happening
to me,
Toshi Hoo (32:09):
exactly. And I think you asked about my history if,
TF, it's interesting. And when Ijoined in 2016 I also look back
to that as a time where therewas a bit of a shift, I think,
about the general public'sperception of is technology
bringing us to some sort ofutopian world. Right until that
point, it was like, oh, socialmedia, it's so fun. We're all
connected. And it was in thosefollowing years that we started
(32:29):
really seeing examples, data andreally analysis, saying, like,
maybe technology, maybe theinternet, is not great in all
ways. It's spreading massivedistant information. It's
causing overthrows of democracyin some places, right? It's a
mixed bag. And I think, youknow, that's really part of what
we try to do that city for thefuture, is to really kind of
expand our range of the types offutures. We don't when we do
(32:50):
forecasts. It's always importantnot just to consider our future,
but a range of futures. And Ithink this is a little bit of
where I diverge from rays. Rayhas a very, kind of singular
view of like, okay, we're thisis where we're going, and that's
exactly what's going to happen.
And the reality, I think, isthat the future is much more
what I call multiversal meaning,not only are there multiple
possibilities, but there's,we're all going to experience
(33:12):
that future from a multipleperspectives, right? Sure, and
those are all very different. SoI think we need to have, I mean,
I always really push for more ofkind of a pluralistic view of
like how we think about thefuture, and really try to dispel
the myth of kind of thefuturists and the ivory tower
who tell you what's going tohappen, and more that we all
need to develop our capability,individually and as collectives,
(33:34):
to the ability to kind of havecohesive and valuable
conversations about the mostimportantly, the futures that
could happen, and what are thepreferable futures we're
actually trying to create. Well,
Matt Kirchner (33:46):
and one of the things you referenced earlier
was the idea that you're notnecessarily trying to predict
the future, that nobody canpredict the future. I read a
book probably 20 years ago andsat through a series of lectures
on this idea of what they calledscenario planning. And the whole
idea was that the goal isn't tofigure out exactly what's going
to happen and get there beforeeverybody else or make sure
you're prepared for that versionof the future. Is to say, okay,
(34:09):
these are the four or fivethings that are most likely to
happen. And if I start to seethe world going in one direction
or another, and you could thinkabout that as broadly as all of
technology, or as narrowly asmaybe things that would affect
an individual business, thenI've got a plan for where to go.
And it sounds like yourphilosophy is almost the same
thing. Little
Toshi Hoo (34:28):
bit. I would take it even further to say that, you
know, I mean, we do createforecasts at the institute that
is sure, our bread and butterforecast scenarios. So forecasts
are descriptions of the future.
We like to say they're plausiblebut provocative statements about
the future that help us makebetter decisions today. Right?
It's not enough just to saysomething of the future. It's
got to be able to be somethingthat we can facilitate and use
(34:49):
today, and then scenarios aremore stories about those
futures. So how do we imaginewhat it would be like to inhabit
those as an individual? Oroperate within those future,
future forecasts as anorganization, it's through
thinking through not just theforecast. You're never going to
come to an iftf event and justread a forecast and be like,
Okay, I got I know what's goingto happen now, right? The point
(35:11):
of the actual kind of productthat I like to say is not just
even the things that couldhappen, but most importantly, as
you run through your ownscenarios around you as an
individual or you as anorganization, that you learn
about your organization, youstart to see yourself and your
organization in ways youcouldn't see before. And it's I
even caution people from thisterm future proofing, because
(35:35):
also that assumes, like, Okay, Iknow probably what's going to
happen. I think if you assumethat you know what's going to
happen, you're going to misswhat's actually happening. Yeah,
that's a really good point. Weare in a world that's
increasingly volatile andunpredictable and chaotic. It's
comforting to think, Okay, Iknow what's going to happen, or
I'm future proofed, and peoplelove that, but the comfort is
(35:57):
actually lulling us into not payattention to what's happening.
Now, in our practice, I kind ofalluded earlier, when we run a
forecast, we won what's calledthe Four alternative scenario
forecasts, which means you don'tjust run a forecast, you play
that forecast out. So forexample, if my forecast is in
five years, we are going to havemore virtual humans than real
(36:17):
humans. And then you play thatscenario out in four different
kinds of archetypes. This is aframework created by Jim data,
who is one of the grandfathersof Strategic Foresight at
University of Hawaii. And thoseare growth scenario, collapse
scenario, a constraint scenarioand a transform scenario. And
it's through running thosedifferent kinds of scenarios on
(36:39):
the assumptions that yourforecast has, that you start to
understand the dynamics of whatcould happen. So if this
happens, that might happen, andthat's going to help you better
be prepared and be what we callfuture ready, not future proof,
but increase your futurereadiness, so that you're able
to kind of re contextualize yourown model of who you are, and
(36:59):
that with the options you havein these new operational
environments, which is,
Matt Kirchner (37:04):
I mean, it's just fascinating to think about that
and to think, what were thefour, by the way, it was
transformative, collapse, growthand constraint. Yeah, so that's
an interesting way to look atthe world, without a doubt, all
the incredible work that you'redoing, at iftf, I've got to
believe you've got, like, onegreat success story of, we have
this nonprofit where there's areason that we exist that's
(37:26):
obviously beyond just generatingcash flow and generating, you
know, economic results. It'swe're doing this great work. You
have a story that might resonatein terms of where you really
help change a life or change amarket space, or do something
that you felt really good about.
I could give you a bunch ofdifferent examples
Toshi Hoo (37:44):
about kind of the work that we've done with big
companies or governmentagencies. One of the challenges
is, when you're doing 10 yearforecasts, you don't often takes
a while how to hear back frompeople if those are, you know,
helpful or useful, but we do.
Matt Kirchner (37:57):
That's the benefit too, is that if you're
wrong, nobody remembers by thetime 10 years goes by, exactly
Toshi Hoo (38:02):
I could talk about projects, and I'm happy to but
to be honest, in my work, thetimes where I feel most
successful with what I'm doingis after I give a keynote, for
example, giving a talk aboutgenerative AI. And when I come
to somebody and they tell me,thank you, you explain that to
me in a way that I could finallyunderstand. It awesome to me.
The future is less going to becreated by like, one great
(38:23):
entrepreneur that's going to bethe one that changes the world,
and more about the kind ofcollective understanding and
literacies that we're buildingacross all of our different
layers of society and differenttypes of folks. Absolutely.
Yeah, I'm happy to talk aboutproject as well, but in terms
of, like, building fitness,that's a perfect
Matt Kirchner (38:40):
example. I think it was in my life's book. And I
know you interviewed, EthanMalik at one point, where he
talks about that being one ofthe great gifts is the ability
to if and if it's not him, it'ssomebody else that I read
recently, but it talks about theability to take really complex
concepts and boil them down intosomething that the average
person, or somebody who isn't asubject matter expert in that
(39:02):
area, can understand andcomprehend and understand the
impact. And I would say youdefinitely have that gift having
seen you speak myself. So that'sa great example of how the work
that you're doing is helping tochange lives. I want to give you
one more opportunity to talkabout something that changed the
life, and that's your own life.
One of the questions we love toask every one of our guests here
on The TechEd Podcast Toshi, isto ask someone to go back in
(39:22):
time. You know, you go back intime to that 15 year old version
of yourself. You know, yourdad's working for Ray Kurzweil,
which had to be just anincredible experience to grow up
in a home like that, before youhad all the success in film,
before all you had all thesuccess in media, and then the
incredible technology stuff thatyou're doing these days, if you
could go back and give thatyoung man one piece of advice,
Toshi, what would that be?
Toshi Hoo (39:44):
That's a great question. As a futurist, I'm
often thinking about the futureand not right about the past.
You know, if I were to go backto my 15 year old self, I would
say, be your natural, curiousself. It's going to take you
places you could never imagine.
You'll go absolutely and it didyou. Know, I actually like to
say that my main product that Iproduce for the world is is
wonder, yeah, meaning motivatedcuriosity. And I think that's
(40:06):
the most transformative thingyou can do. And it's and it's
not just kind of like a playful,you know, fun feeling. It's more
helping people imagine thatmaybe the world is even larger
than they imagined. No question,it is, leads into asking
questions and growth. And Ithink that's really how the
world is really transformed. Sothat's how I got to where I'm
(40:28):
at. So if I were to say my 15year old self, I'd say, just
keep being curious. Don't try toguess where you're gonna go.
That's right,
Matt Kirchner (40:36):
never really, yeah, no, and that's so
fascinating. And I've been onthis actually, this kick about
curiosity this year, and it'sbeen influenced by guests we've
had on the podcast. So I'll giveyou just three quick examples.
Mike Bigley, who is thesuperintendent of a school
district in western Wisconsin,just created a whole emerging
technologies lab for hisstudents and all kinds of really
cool applied AI experiences forthose students when he's looking
(41:00):
for educators to be part ofleading change like that. He
said the number one personalitytrait he looks for is curiosity.
Then we had Todd wanick, who'sthe CEO of Ashley Furniture,
good friend of ours, largestfurniture manufacturer in the
world, of course. And we askedTodd, when you look for people
to lead your AI transformation?
I said, What do you look for?
And he said, Curiosity was thefirst thing that he'd look for.
(41:22):
And then we had Barbara humpton,who's the CEO of Siemens, huge
company, of course, Siemens,USA, $20 billion in revenue,
45,000 employees. Give or take.
One of the things that Barbarahumpton said is, if you have
curiosity and initiative, theworld is yours, which I thought
was really, really interestingway of looking at the world in
the fact that here you are withall the cool experiences that
(41:43):
you've had. Nobody knows whattheir career journey is going to
look like. You've had anincredible career, no doubt more
to come. But for you to say,Hey, I would remind my 15 year
old self to be curious is justone more data point on this
whole journey of curiosity thatI've been on here over the
course of 2024
Toshi Hoo (42:00):
a lot of people come up to me right now and say, What
should I tell my 10 year oldchild? Now I actually think the
future belongs to the curiouspeople. Right? We're in a period
of time where so much is beingdisrupted, and that's scary, but
it's also an incredible time fornew things to happen, and it's
the curious ones that are goingto really create and explore
those The future
Matt Kirchner (42:19):
belongs to the curious people. I can't thank
the curious Toshi, who enoughfor joining us here on The
TechEd Podcast. It's been aphenomenal conversation. We
reference some things I know ouraudience is going to want to
check out. You will find thoseand all of our show notes at
TechEd podcast.com/who that is.
TechEd podcast.com/h O, O, needto remind our audience as well
(42:42):
to check us out on social media.
As you know, we are on LinkedIn,we're on Facebook, we are on
Instagram, we are on x, we areeverywhere you would ever want
to look for your social media.
So while you're there, reach outsay hello. We would love to hear
from you, and we would love tosee you again next week on The
TechEd Podcast, thanks so muchfor being with us. You.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com