April 18, 2026 • 35 mins
In this episode of the News Items Podcast, John Ellis speaks with venture capitalist Daniel Adamson about investing at the cutting edge of global innovation. Adamson explains his “network model,” partnering with elite investors worldwide to access breakthroughs across AI, biotech, robotics, and even fashion. The conversation cuts through AI hype, arguing for a more nuanced view of its strengths, particularly in solving complex, data-heavy, rule-based problems like drug discovery. Adamson highlights transformative opportunities, from automated science labs to humanitarian applications, while warning of disruption across sectors like infrastructure and real estate. Ultimately, the episode explores how rapid technological change is reshaping markets, portfolios, and even our understanding of what it means to be human.
Daniel Adamson operates as a network-driven investor rather than a traditional venture capitalist. Instead of selecting individual startups, he invests in leading venture capital firms and partners with major global institutions such as pensions and sovereign funds. His approach centers on identifying top investors across industries like biotech, fintech, and energy, then aligning with them by taking stakes in their firms and co-investing in their strongest opportunities.
This model provides access to high-quality deals, diversified exposure, and a steady flow of insight from experienced managers. By connecting capital with specialized expertise, it leverages collective knowledge rather than relying on a single viewpoint.
The advantage is early access to innovation and efficient deployment of capital. Rather than trying to outperform specialists directly, his role is to build and manage a system that consistently directs investment toward the most promising opportunities worldwide.
Hosted by John Ellis
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:00):
Hello and welcome back to the News Items Podcast.
I'm John Ellis.
I'm the founder and editor oftwo Substack newsletters.
One is called News Items, theother is called Political News
Items.
You can find them both atnews-items.com.
Our guest today is DanielAdamson.
Dan is a venture capitalist andthe founder of Collective
(00:22):
Global, a firm that works with adozen of the world's largest,
most sophisticated investorsacross Scandinavia, Australia,
California, and beyond.
The firm buys stakes in topventure managers at the general
partner management companylevel, and then co-invests
alongside them in their bestdeals.
Dan's firm covers industriesfrom AI and biotech to financial
(00:46):
technology, energy, andinterestingly, fashion.
I wouldn't have guessed that.
Dan's educational background isextraordinary.
He's a Sumacum Laddie graduateof Yale College, a Marshall
Scholar at Oxford University,and a graduate of the Yale Law
School.
Dan, I'm going to overlook thefact that you weren't the editor
of the Yale Law Review and askyou: once you graduated from
(01:08):
Yale College, did you go to workat a law firm or did you go
straight into the financialworld?
SPEAKER_01 (01:13):
You know, I quickly realized that I was not cut out
to be a lawyer.
So I actually graduated from thelaw school, but spent my third
year working for a venturecapital firm in New York City
and just went back to New Havento take my exams.
So you could look at that as anincredibly dumb decision in that
I spent all that tuition moneyand what did I really get for
(01:34):
it?
On the flip side, I was already26, been in school for a long
time, was itching to get outinto the world, knew what I
wanted to do, and New York Citywas just a little over an hour
by train.
SPEAKER_00 (01:45):
Yeah.
I would say that the besttraining for anything is to go
to law school because it teachesyou how to think, and therefore
you're going to be better atanything else you might do.
That proven true for you?
SPEAKER_01 (01:57):
You know, in my case, boy, those those other
kids at the law school areawfully smart.
And so I think it taught me thatI was not going to be the best
at anything.
I've got three kids, ages uh 11,12, and 13.
My 12-year-old is realizing atthe moment that he is not going
to play for the New YorkYankees.
Um and I I kind of had theparallel realization about every
(02:19):
professional track uh by meetingthe quality of my peers at at
Yale Law School.
And so I made a pivotal decisionthen, which has influenced my
career ever since, which was ifI'm not going to be the best at
anything in particular, let meat least try to be the best at
partnering and go find the bestpeople in each area.
And you know, combine that withenthusiasm as a kind of lifelong
(02:43):
learner, enthusiasm forinnovation, joy of making
investments, and well, you get acompany like the one you
described.
SPEAKER_00 (02:51):
So you describe your business employing a network
model.
Can you tell us exactly?
I mean, I think of you know,there are there are hedge funds
and then there are fund hedgefund of funds, I guess you would
call them.
And your model seems to beloosely like that in that you
don't have a venture fund, youinvest in the best venture funds
(03:12):
and then co-invest alongsidethem.
Is that is that more or lesscorrect?
SPEAKER_01 (03:16):
Yeah.
Um the the way I describe it inthe simplest terms is just that
we are building a better bridgebetween capital and innovation.
It seems to me that if there'sone thing, and there may only be
one thing that everybody in theworld can agree on right now, it
is that the pace of innovationis astonishing.
I mean, pedal to the metal, sortof back of your head hits the
(03:39):
headrest in the car, right?
That level of pace.
And if you're gonna invest inthat way, you are at the cutting
edge.
And by definition, you're notgonna understand everything that
about everything.
Best thing you can do is topartner.
So rather than model ourselvesin a conventional way where
we're just trying to have asmany investors as we can, do as
(04:02):
many deals as we can and hopethat they all go well, hundreds
of investors, hundreds of deals.
We've taken an oppositeapproach, really, where we we
work with deep-pocketed,sophisticated investors,
intentionally global, and limitthe number of them so that we
can really mind meld with them.
And then on the manager side, wetry to find the best managers in
(04:24):
the world, groups like GeneralCatalyst, others that that are
household names, and we buystakes in their top coat.
So we are a minority owner oftheir business, not just an
investor in their funds or aco-investor alongside them,
those, though we do those thingsas well.
And what you get is a kind ofnetwork.
I mean, everybody's talkingthese days.
(04:45):
We're obsessed with the idea oftaking nodes, connecting them
together, and buildingartificial intelligence.
I'm sure we're gonna talk aboutthat today.
I hope we have one of the firstactually interesting
conversations about AI in sometime because it feels like it's
all anyone wants to talk about.
And I hear the same things allthe time.
So we'll try to say somedifferent things.
But one thing that um we dodifferently is we're building a
(05:10):
network, a kind of neuralnetwork, if you will.
But rather than those nodesbeing neurons or chips, they are
organizations and people who areexpert in what they do.
And they're from all over theworld, and they know areas like
energy and you mentionedfashion, right?
We've got a consumer businessthat we back.
We we we look at biotech, welook at fintech, and we work
(05:33):
together not just in the waythat people chat when they're at
Davos, you know, everyone'spuffing out their their peacock
feathers and and trying to bethe biggest organization in
town.
Our strategy is different.
We actually are in businesstogether.
So we're not just meeting at acocktail party.
We own each other's businesses.
(05:55):
That creates a permanentconnective tissue in this sort
of neural network.
And then because the network isbuilt out of people, not chips
or neurons, we can actually gettogether.
And that's when it comes alive.
So every year we host ourcompany, Collective Global.
It's actually hosted by one ofour asset owner, pension, or
sovereign partners.
Last year it was in Stockholm,and we had our major event at
(06:19):
the hall where they give out theNobel Prizes.
This year it'll be in Nice,France, next year in Sydney,
Australia.
And that's when this kind of, Iwon't call it it's not an
artificial intelligence, butthis kind of organizational mind
comes together.
And it helps us all becomebetter thinkers, sharper
thinkers about what's happening,especially in innovation.
SPEAKER_00 (06:40):
So we have to talk about AI.
The three topics that I thinktop of mind for investors,
obviously, AI is one of them.
Robotics, I saw it was describedas the largest addressable
market in human history.
So there's robotics, and thenthere's another kind of code,
which is genetic code.
You're active in all three ofthese areas, but let's start
(07:02):
with artificial intelligence.
We have, you know, we the thediscussion is doomsday or
unlimited promise.
You argue that that's not theway to look at it.
Tell us why.
SPEAKER_01 (07:14):
Yeah, I mean, first of all, I I think if you ask 12
experts on AI what they think,you'll get what's the joke, 13
different opinions.
Right.
And nobody really knows becausethe pace of change is just so
incredibly fast.
I will make a promise here thateven though I'm not a STEM
person per se, I was aphilosophy grad student and then
(07:36):
a lawyer, I think that actuallycan help us cut through some
jargon and talk about uh AI in aslightly different way.
So instead of just throwingaround words like deep learning,
foundational models,reinforcement learning, genetic
token, is it a large languagemodel?
Oh, then one day it's supposedto be what about small language
(07:58):
models and you know, uhmultimodal.
And my favorite word ishallucination.
I mean, what are we really,really talking about here?
So I think that rather thanhaving a simplistic doomsday,
you know, versus utopianRorschach test where you just
say AI and you see whetherpeople's blood pressure goes up
(08:20):
or down, it is possible to havea nuanced view of what AI is
good at.
We are spending roughly 1% ofgross world product, a trillion
plus a year, on the stack ofwhat is AI, from energy through
to data centers, through tochips, through to foundational
(08:41):
models, through to apps, throughto agents, ultimately, in some
cases, as you mentioned, throughto robots and actual
experiments, we're spending aton of money on that.
You would have to go back tomore than 100 years ago.
I think some historical contextis helpful, if you'll forgive
me.
Like when was the last time wedid that as a private sector?
(09:01):
I know we spent 3% of GDP onputting a man on the moon, but
when was the last time that theworld just organically decided
through the minds of investorsand business people to put this
much into one coherent area oftransformation?
I would say, you know, the callit 30 to 50 years where we
invested in railroads,electrification, and
(09:24):
agricultural mechanization,which were obviously three
linked technologies.
The transformation that camefrom spending 1% of gross world
product in those three areasover the call it roughly second
half of the 19th century wasincredible.
Imagine grabbing somebody fromthe year 1800 in America when
(09:47):
90% of people worked on farms,and telling them that in 2026 we
were going to have 2% of peopleworking on farms, and
unemployment would be less than5%.
And then asking them, what'severybody gonna do for a living?
I don't think, I mean, theycertainly wouldn't come up with
management consultant or podcasthost or electrician, right?
(10:09):
Those were not on the menu atthe time.
So I just say that because theamount that we don't know
compared to the amount that wedo know is so vast.
What we don't know is an ocean,what we do know is a drop of
water.
We can say some things aboutthat drop of water, and I'm, you
know, I think we can take outthe microscopes and say some
interesting things about it.
(10:30):
There's a lot that we stilldon't know.
So jargon's not going to help.
It just sells news stories andgets people anxious and maybe
drives up valuations on certaininvestments.
But there are coherent thingsthat we can say about AI.
Certainly we can expecttransformation that is at least
equivalent to the transformationthat we saw coming out of that
(10:53):
tri-pronged era of innovationwith railroads, electrification,
and agri agriculturalmechanization that I talked
about, in which the world becamethe one that we know today.
And the difference here is it'sgoing to happen much faster.
It's already happening muchfaster.
And it's happening in ways thatwill have some predictable
consequences and some second andthird order consequences that
(11:14):
are very difficult to predict.
SPEAKER_00 (11:15):
I saw a brief clip this morning on YouTube.
It was one of the leading AIalgorithm writers, I guess you
would call them.
And he was saying that herecently attended an AI
conference and was himselfastonished by what was projected
to happen, not in the next sixyears, but in the next six
(11:36):
months?
I wonder, as somebody whoinvests in funds that invest in
AI and that, you know, co-investon a number of those deals.
Is the pace of change AI soextraordinary that it sort of
freezes you, if you will, fromfrom investing?
Do you have a pretty good ideaof what you want to invest in
(11:56):
and there's a rationale behindthat?
SPEAKER_01 (11:58):
Aaron Ross Powell It's a little bit of both.
I think you need to be humble inrecognizing that we don't know a
lot, like I just said.
But I also said that, and I madea promise that we'd try to have
a conversation that was coherenthere about AI, which is rare
these days.
And let's start with a theory ofwhat kind of intelligence we're
(12:19):
building.
I think everyone's talking aboutwe're building AI, and they they
feel like that's enough.
Oh, I just said we're buildingartificial intelligence, as if
that's a sufficient description.
It's not a helpful description.
What type of intelligence is it?
What is it good at?
What is it not good at?
When is it going to be good atthe different things that it
could help us with in society?
(12:40):
So let's try to unpack that.
When did we all first hear aboutAI?
Well, I think it's when DeepMind beat Kasparov at chess.
Right.
I forget the year, but that youknow chess was the first
example.
Then AI had a win again in amore complex game, a Go, where
beat the world champ Lee Sadal.
(13:02):
If you put those two data pointson a chart, and now I'm going to
put a third one up there becauseI've just I can report an a very
interesting breakthrough, acompany that I serve on the
board of called Lila Sciences,where they have crafted an mRNA
molecule that in vivo has anorder of magnitude better
(13:23):
results in fighting certaincancers in the human body than
Big Pharma has come up with.
And they did it with AI.
What do those three data pointshave in common?
I think they're examples ofwhere AI is flourishing as a
type of intelligence.
They have in common that theyare all based on rules.
Chess has rules, Go has rules,DNA, mRNA have has rules.
(13:47):
Secondly, you've got immenseamounts of data, just brute
force, more data than a humanbeing can hold in its mind at
any given time.
So the the analogy I like to useis if you've got 10 light
switches on a wall, they can bein a thousand and twenty-four
different arrangements.
If you've got a hundred lightswitches on a wall, there's more
(14:09):
potential arrangements thanthere are atoms in the universe.
Now think of a chess boardrules.
You got huge amounts ofpotential formations of all the
pieces, and you're quickly, youknow, at more than the number of
arrangements than there areatoms in the universe.
Go was supposed to be impossiblefor AI because it was an order
(14:30):
of magnitude or more complexthan chess.
Well, mRNA, of course, exists,and RNA exists because it's how
nature chose to encodeinformation.
And the combinatorial landscape,to get fancy about our language
here for a second, that iscreated through four base pairs
interacting in almost an endlessvariety of ways and then coding
(14:53):
proteins, which can then expressin an almost endless variety of
ways, is shocking.
It's important, and it'ssomething that humans are just
not very good at doing.
We have made great progress.
I don't mean to suggest that thelast 450 years of science
post-Bacon and Descartes andothers in the early 1600s hasn't
(15:16):
been incredible.
But the pace at which we canconquer this particular type of
problem, one with rules, onewith vast amounts of data, and
one with huge combinatoriallandscapes, is is now newly
solvable.
So what's in that space andwhat's not in that space?
Well, I would argue thatcertainly this is going to be
the century of biology, right?
(15:37):
Biology happens to be reallycomplicated.
And so it fits what I justdescribed to a T.
So we've done a lot of investingin enabling AI to access
biological systems.
They often do it through robots,because robots can absorb much
more data.
And biological systems areinherently incredibly complex.
I think there are somenon-biological scientific
(16:00):
systems, like how do we replacethe, what is it, seven rare
earths that China's given ustrouble over in terms of their
use as catalysts in differentindustrial and other military
processes?
Well, that's the sort of thingyou need to run experiments on.
And turns out nature iscomplicated.
You can't just think your waythrough it.
You need to actually go run theexperiments, onboard the data,
(16:23):
analyze the data, synthesize thedata, and then run the next best
experiment after that.
So those are some areas where AIcan be helpful.
I'll give you one more justbecause I don't want both of my
examples to be focused onscience.
You know, AI gets a wrap thesedays.
You talk about the dystopianangle, it's going to replace all
of our jobs, et cetera.
(16:44):
There are also some hugehumanitarian crises that don't
get discussed enough in theworld, in my view, where AI
could, if allowed, be incrediblyhelpful.
The UN is a UN statistic.
There are 154 million orphans.
That is a shocking number.
It's 2% of all humans alive thatare under age 18 and don't have
(17:04):
parents.
Um it's something I knowpersonally because my wife and I
have been very active andadopted a kid.
And uh the reason there are somany is well, it's very
complicated and it's it's it'sgeopolitical and it's it's we
just don't live in a fair world.
But one other reason is thatthere's countless languages,
tons of paperwork, and 154million is a big number.
(17:26):
Well, what's AI good at?
Reading and understanding anylanguage, following any legal or
bureaucratic process, andcapturing and processing
enormous amounts of data.
Think about the amount of moneywe've spent on dating apps to
match, you know, Joe and Sue,who live down the street from
another, one another in, youknow, in Urbana or in Columbus
(17:49):
or in Atlanta.
How about we spent a fraction ofthat matching kids with families
that want kids?
I'm not saying that it's such aneasy problem, but it is a
problem that AI could help withbecause it is a large
combinatorial space that isrule-based with a ton of data.
SPEAKER_00 (18:07):
I wanted to talk a little bit about Lila.
I'm just fascinated.
How did it, presumably you'dnever heard of it like three
years ago, now you're investedin it?
How did how did you come to knowabout it?
What convinced you to invest init, et cetera?
SPEAKER_01 (18:22):
Yeah, I certainly hadn't heard of it.
It was in stealth mode.
But one of our manager partners,General Catalyst, who I
mentioned earlier, and we've gotnow six manager partners.
Collective Global, my firm, isis is relatively new.
I've been doing some variant ofthis kind of investing for 25
years, but we decided to focuson innovation in 2023.
(18:42):
I think our timing was good.
And we backed General Catalystalongside McKinsey, JP Morgan,
and Amazon.
We were one of their backers atthe top co level.
And they've been wonderfulpartners.
And turned out that, if Iunderstand it right, the biggest
safe, which is a particular typeof early stage investment that
GC General Catalyst had everdone in its history was within a
(19:06):
company called Lila Sciences.
So immediately my my ears perkup and they say, You want to see
the future of science?
Go to Boston, check out thislab.
And they send me a picture ofBill Gates, you know,
practically fogging up hisglasses.
He's so excited about what he'slooking at as he's in a lab coat
in the Lila Sciences facility.
(19:26):
The most interesting thing aboutthe photo is that nobody is
touching anything.
So if you think about the wayscience has been done for the
last several hundred years,right, since the age of Louis
Pasteur, right?
We had grad students withpipettes or the equivalent,
moving something from Petri dishA to Petri dish B.
Then they spill their coffee onit.
(19:47):
Then, you know, the their theirthesis advisors argue with them
about it.
Then they tenured professorpublishes something on it.
Then they argue with a differenttenured professor about whether
they were right or this othertenured professor was right, and
did they falsify the data?
And it's it's a mess.
Human science is a mess.
It's also the most productivething we've ever figured out how
(20:10):
to do.
So don't I don't I'm I'm notdemeaning science, right?
I'm just telling you, it's amessy business.
It takes a while.
Now picture instead the insideof one of these semiconductor
slash chip manufacturers, youknow, with the clean hazmat
suits and the 30-foot ceilingsin places like Taiwan.
It's just incredible.
(20:31):
Imagine we were applying thatprecision to the biological
sciences, to the materialsciences in other domains.
Well, that's what Lila does.
And so I could appreciate thatthe entire wheel of science
itself needed an update, andthat putting AI with the human
in the loop at the top of thatchain, and then using robots
(20:53):
that could ultimately synthesizetheir own equipment through 3D
printing to improve recursivelythe quality of the experiments
that they're running, run them24-7, they don't spill coffee,
they don't make a mistake aboutwhich petri dish they just put
the sample in, they suck uporders of magnitude, more data,
they process it faster, andbecause they can hold in their
(21:14):
mind at one time a much largercombinatorial space, like the
space of all mRNA to fight aparticular type of cancer or
other disease, they can progressmuch faster.
And so this is one of thoseareas where whether it takes six
months, as you said, John, or oror or six years or even twenty
years, if you're a big assetowner, that's what I call the
(21:37):
sovereign's pensions, insurancecompanies, and others that that
I work for really, that I helpconnect to innovation, it almost
doesn't matter if it's sixmonths from now or six years
from now.
You've got liabilities to yourpensioners that last 10, 15, 20
years, sometimes longer.
If these disruptions through thesciences or in other areas, we
(21:58):
all witnessed the SAS pocalypserecently.
Which we could talk more aboutlater, where private credit got
thrown under the bus.
If these changes are going tohappen in the life cycle of your
liabilities as a pension orsovereign, you need to be
worried about them.
And it is evident to me with,and I don't have a crystal ball,
but I do have call it 12 to 24months advance warning on what's
(22:20):
happening because of theinvestments that we're making
that are in small privately heldcompanies.
This stuff is coming not justfor new discoveries.
It's going to change the other98% of your portfolio in ways
that aren't always positive.
So you need to be eyes up, chinup, thinking hard about
innovation today in a way thatfive years ago you could maybe
(22:43):
justify not having a venturegrowth portfolio.
Today, I don't think any CIO inthe world can hold their head
high unless they reallyunderstand and make an effort to
understand the pace of change.
SPEAKER_00 (22:55):
Aaron Powell Just as an example, the impact of the
change that is coming on otherparts of one's portfolio.
Are there specific sectors thatyou think are most at risk or
specific companies that youthink are most challenged, I
guess is the polite way to putit?
SPEAKER_01 (23:12):
Aaron Powell Well, yes.
I actually think, for what it'sworth, that this recent SaaS
pocalypse, as I called it,software as a service getting
hammered.
Uh and in particular, I think itwas 40% type drops in the listed
private credit asset managerslike Aries and Blackstone and
Blue Owl was an overreaction,personally.
(23:34):
I think a lot of these SaaSbusinesses are good businesses.
They got hit hard because, asyou know, John Claude Code came
out, and all of a sudden peopleare worried, well, maybe I'm not
going to need to buy Salesforceor some other SaaS business.
And if you're an Aries andyou're lending to them and
you've got a duration on yourprivate loan to that company
(23:55):
that that goes into the 2030s,and all of a sudden you're
thinking that the equity mightbe wiped out, and maybe even
they won't be able to cover alltheir debt in 2032.
Well, that starts to be aproblem, and the market, I
think, overreacted, priced thatin quickly.
My particular view, which mightbe wrong, might be right as to
why that was an overreaction, isthat, you know, let's take
(24:16):
ourselves back in time to whenYouTube came out.
When YouTube came out, itfloundered briefly because
everyone thought, why do we needmore video?
There's tons of high-qualityvideo.
You can stream it this way, youcan rent it that way.
Turns out we had a lot lessvideo than we really needed, as
my tw you know, fifth, sixth,and seventh graders' behavior
patterns every day will testifyto.
(24:39):
So maybe we just have a lot lesscode than we really need.
And so this reaction to say,okay, now companies that provide
code, software as a service, aregoing to die, I think has been
overstated.
So that happens to be one I'mnot worried about, but you are
not as worried about.
You asked me about some areasthat I am worried about.
So here's a big one.
Like what happens whenself-driving cars actually turn
(25:02):
on?
I mean, we we all believe thatthat's gonna happen, right,
John?
I mean, it could be some of it'salready happening, but i i i i
is it gonna be five years, tenyears, fifteen years when this
horrible corridor that you and Ideal with between Fairfield
County and New York City startsto flow like like water instead
of like molasses because it'sself-driving cars.
(25:24):
Well, maybe your publicinfrastructure bets, which you
thought were the safest part ofyour institutional portfolio,
your light rail investments incities like LA or Toronto, which
were built around the car thatyou thought were safe, maybe
they're not so safe.
I'll give you another example.
Your commercial real estateplays in the life sciences,
(25:45):
where you thought, oh, well, ifI if I have if I have wet lab
space near researchuniversities, boy, that's
probably going to be a good 10to 20 year investment.
Well, I've just given you apicture of what science might
look like, will look like in myview, in five, 10, 15 years.
And turns out robots don't likenine-foot ceilings or 10-foot
ceilings.
(26:05):
They need 20-foot ceilings,depending on exactly what's
happening in the space.
So all of a sudden, what seemedlike a safe bet isn't.
And to me, that's the it's boththe returns you can get from
innovation, but also theinsights that you need to have,
the kind of early warningsystem, if you will, that's
gonna predict the seismicactivity in other parts of your
(26:27):
portfolio.
I don't see people talking aboutthat.
Just like I don't see peopletalking about the type of
intelligence that AI is actuallysmart at.
I mean, there's tons of thingsAI is not good at, by the way.
Like I wouldn't pick it to be mysports agent if I were a good
athlete, right?
It's not gonna negotiate a dealand understand all the moving
(26:48):
parts in a complex humansituation.
But it's very good at thesekinds of combinatorial space
thinking that I talked about.
Similarly, there are industriesthat won't be impacted.
I think live entertainment isfairly AI proof.
My kid will still be able toroot for the Yankees in 20
years, even if he won't be ableto play for them.
But there are lots of areas thatwill be impacted.
SPEAKER_00 (27:08):
Aaron Ross Powell There's a group out in
California called Rethink X, andthey did a paper on
transportation as a service.
And so in this model that theybuilt, the cars are electric,
they're self-driving.
You might own one, but you onlyuse it an hour a day, so you
could rent it out three or fourhours a day.
The electric car has 128 movingparts.
(27:32):
Uh combustion engine has, youknow, a thousand moving parts.
So maintenance of that car is,let's put it this way, much less
than it would be for acombustion engine car.
Because it's self-driving andpresumably safer than current
situation, insurance costs willgo low.
That kind of disruption ifyou're looking at
(27:52):
transportation, is thatsomething that that you're
thinking, okay, this is reallygoing to be totally different,
and how do we invest in thisspace?
Or I guess you know you'd justsay transportation's too c you
know too too outside of ourspace, so we'll leave it be.
SPEAKER_01 (28:10):
Well, we we may not be, uh part from our investments
in self-driving cars, which wehave, we may not be focused on
big public infrastructure typeinvestments like the light rail
or or or commuter rail that I Italked about a moment ago.
But our pensions and sovereignsare, and those are their biggest
bets sometimes.
Multi, multi-billion dollarinfrastructure plays that they
(28:31):
thought of, oh, well, this willgive me a five, six, seven
percent yield consistently forthe next 25 years.
Losing steam is a much biggerdeal than missing out on the
next trillion dollar startup forthem.
unknown (28:42):
Right.
SPEAKER_01 (28:43):
Right?
Because they're gonna own asmall percentage of that
startup, but they already own abig percentage of those
infrastructure plays.
And I think what we need to getcomfortable with, and we started
the conversation by talkingabout the sheer pace of change.
Talk about electric cars, right?
You you press the pedal to themetal, so to speak, in a you
know, an electric car and thetorque throws your head against
(29:03):
the headrest.
That's the pace of technologicalchange today.
And outside of wartime in humanhistory, we're just not used to
thinking about the future beingso dramatically different than
the present in a fairly shortperiod of time, five to ten
years, maybe less.
So we've got to be on the ballsof our feet, the tips of our
(29:26):
toes as investors.
And even that will just give usan edge, not uh not a right to
win.
So, but but an edge is is betterthan than sticking your head in
the sand in this moment.
So I'm hoping that more CIOswill pay attention to the
innovation economy and thinkabout these second and third
order consequences for what theythought of as the quote unquote
(29:49):
safe part of their book.
You know, the the deeperquestions here that that you
know maybe maybe we end withare, you know, how is this gonna
even change the role of a CIO,of a decision maker as an
investor?
You know, it's in in every erawe humans have thought of
ourselves as the protagonists.
(30:10):
We're the ones driving history.
And we're now entering an era inwhich that is, I would say at
best uncertain.
And to me, again, you know, youcan look at that and and just
get depressed, or you can lookat that and get really excited.
And probably you can look at itand feel both ways at the same
time.
But I'll tell you what makes meexcited about it is that you
(30:30):
know, we don't really have a atradition that helps us think
about what humans can offer ifit's not being the protagonist,
if it's not being the smartest,the toolmaker, the species above
the others.
And and if we are no longer thesmartest, if we are no longer
the toolmakers, if we are nolonger the species above the
others, because we we gave birthto it in the form of AI, it's
(30:54):
going to force us to ask allthese really interesting
questions about what does, ifanything, make us special?
And I think I get excited about,you know, we started this
conversation talking about my myfamily, and you know, we've
grown it through adoption.
We I've I've also I think thethe greatest single act of
goodwill I've ever done is thatI'm originally from Atlanta and
(31:15):
I've allowed my 12-year-old sonto be a Yankees fan.
Just think about that for aminute.
Right.
What that's real personalsacrifice that involves.
But so so like we've got toreimagine what it means to be a
human in the context of notbeing the smartest, of not
necessarily being the mostefficient, most productive
protagonist uh on earth.
(31:37):
And I think that's gonna leadto, I hope, a revival of of
certain types of arts andculture that I know we both care
deeply about.
And uh I'm excited.
I'm excited both for theeconomic side of things and for
this non-economic side.
SPEAKER_00 (31:50):
Aaron Powell The question people always ask
people like you is you're you'reobviously plugged into a network
of extraordinarily intelligentpeople, but you also read a lot.
What do you read?
SPEAKER_01 (32:02):
As you know, John, we trade books.
So I'm working my way down thebook list you sent me.
But ever since I discoveredaudiobooks, and I can read them,
read them as I'm falling asleep,I read them while I wake up in
the morning, I read them whenI'm on the train, I read them
when I'm between meetings.
I now get a chance to read morethan I have since college.
I felt really guilty about thatfor a while.
I thought it was somehowcheating.
(32:23):
And then I remembered that,well, for most of the call it
30,000 years that human beingshave been telling stories, give
or take, it's been oral, notwritten.
So I thought, okay, I could Icould Homer was mostly, you
know, an oral tradition.
Ultimately somebody wrote itdown, kept it in a in a medieval
monastery so that it could bepassed down to us.
(32:45):
But so once I got over thatshame of being an audiobook guy,
I read pretty much everything,John, and more nonfiction.
I I feel like I've I've come tothe end of what I can read in
physics and uh life sciencesbecause I'm not smart enough to
read the scientific papers, andall of the popular literature
has just become repetitive forme at this point.
(33:06):
So I read a lot of history and Ijust try to find stuff I've
never heard about.
So I just finished a wonderfulbook called Sea Peoples about
the history of Polynesia, andthen I went there with my
family.
It was so great.
So I'm just looking for areas Idon't know about, and then I'll
click on that.
SPEAKER_00 (33:23):
Aaron Powell We've reached, uh I think uh we've
taken enough of your time here,but I wanted to ask you one last
question, which is you'reinvested in AI, you're invested
in companies that do code acrossartificial intelligence,
robotics, genomics, but you'realso invested in fashion.
So the question is why?
SPEAKER_01 (33:43):
Aaron Powell You know, it's funny.
The investment in my portfoliothat my kids are the most
excited about is not LilaSciences, all respect to them.
It's not vast data, it's not anyof these big, you know, unicorns
in the AI space.
It's the fact that we ownindirectly a bit of Justin
Bieber's fashion company.
And I asked my 12-year-old, aspart of my diligence for the
investment, John, I said, Whatdo you think?
(34:06):
Justin Bieber, fashion.
He said, Dad, Biebes is back.
Beeps is back.
Beebs is back.
And so I thought, okay, I'lltake that into consideration.
And then he asked me a question,which I thought was, if I can
brag on my own son for a minute,I feel like that's allowed.
He asked me a question that Ithought, okay, this kid could
have a future in venturecapital.
(34:26):
So he said to me, Dad, does heown the company?
Or is it like one of thoseJordan Nike things where he just
owns the brand?
Well, great question.
He owns the company.
And so then my son said, Youbetter, you better invest.
Piebes is back.
He owns the company.
Go for it.
So we did.
So anyway, I mean, it just tocome to your question,
innovation, it's an it's an easytime to be reductivist about
(34:49):
innovation and just think allinnovation is AI.
It's not.
Like we're also putting nuclearinto space and we're investing
in Justin Pieper's fashioncompany, right?
Like it's it's the one thing AIis actually not doing is
originating truly new content atthe frontier of any field yet.
(35:09):
And that includes fashion.
And so if we can be there withthe best partners in the world
who actually understand thesethings that I certainly don't,
because I started this by sayingI'm not the best in the world at
anything, but I try to be thebest partner.
If we can be there with them,sourcing and then executing on,
and then hopefully adding valueto the best investments in
(35:30):
innovation, it'll span the wholewaterfront from AI to fashion.
SPEAKER_00 (35:35):
I think we've found the uh headline for this podcast
The Beebs is back.
Uh Dan, thank you very much foryour time today.
Uh we look forward to speakingto you again in the future.
Pleasure, John.
Thanks so much.
Hello and welcome back to the News Items Podcast.
I'm John Ellis.
I'm the founder and editor oftwo Substack newsletters.
One is called News Items, theother is called Political News
Items.
You can find them both atnews-items.com.
Our guest today is DanielAdamson.
Dan is a venture capitalist andthe founder of Collective
(00:22):
Global, a firm that works with adozen of the world's largest,
most sophisticated investorsacross Scandinavia, Australia,
California, and beyond.
The firm buys stakes in topventure managers at the general
partner management companylevel, and then co-invests
alongside them in their bestdeals.
Dan's firm covers industriesfrom AI and biotech to financial
(00:46):
technology, energy, andinterestingly, fashion.
I wouldn't have guessed that.
Dan's educational background isextraordinary.
He's a Sumacum Laddie graduateof Yale College, a Marshall
Scholar at Oxford University,and a graduate of the Yale Law
School.
Dan, I'm going to overlook thefact that you weren't the editor
of the Yale Law Review and askyou: once you graduated from
(01:08):
Yale College, did you go to workat a law firm or did you go
straight into the financialworld?
SPEAKER_01 (01:13):
You know, I quickly realized that I was not cut out
to be a lawyer.
So I actually graduated from thelaw school, but spent my third
year working for a venturecapital firm in New York City
and just went back to New Havento take my exams.
So you could look at that as anincredibly dumb decision in that
I spent all that tuition moneyand what did I really get for
(01:34):
it?
On the flip side, I was already26, been in school for a long
time, was itching to get outinto the world, knew what I
wanted to do, and New York Citywas just a little over an hour
by train.
SPEAKER_00 (01:45):
Yeah.
I would say that the besttraining for anything is to go
to law school because it teachesyou how to think, and therefore
you're going to be better atanything else you might do.
That proven true for you?
SPEAKER_01 (01:57):
You know, in my case, boy, those those other
kids at the law school areawfully smart.
And so I think it taught me thatI was not going to be the best
at anything.
I've got three kids, ages uh 11,12, and 13.
My 12-year-old is realizing atthe moment that he is not going
to play for the New YorkYankees.
Um and I I kind of had theparallel realization about every
(02:19):
professional track uh by meetingthe quality of my peers at at
Yale Law School.
And so I made a pivotal decisionthen, which has influenced my
career ever since, which was ifI'm not going to be the best at
anything in particular, let meat least try to be the best at
partnering and go find the bestpeople in each area.
And you know, combine that withenthusiasm as a kind of lifelong
(02:43):
learner, enthusiasm forinnovation, joy of making
investments, and well, you get acompany like the one you
described.
SPEAKER_00 (02:51):
So you describe your business employing a network
model.
Can you tell us exactly?
I mean, I think of you know,there are there are hedge funds
and then there are fund hedgefund of funds, I guess you would
call them.
And your model seems to beloosely like that in that you
don't have a venture fund, youinvest in the best venture funds
(03:12):
and then co-invest alongsidethem.
Is that is that more or lesscorrect?
SPEAKER_01 (03:16):
Yeah.
Um the the way I describe it inthe simplest terms is just that
we are building a better bridgebetween capital and innovation.
It seems to me that if there'sone thing, and there may only be
one thing that everybody in theworld can agree on right now, it
is that the pace of innovationis astonishing.
I mean, pedal to the metal, sortof back of your head hits the
(03:39):
headrest in the car, right?
That level of pace.
And if you're gonna invest inthat way, you are at the cutting
edge.
And by definition, you're notgonna understand everything that
about everything.
Best thing you can do is topartner.
So rather than model ourselvesin a conventional way where
we're just trying to have asmany investors as we can, do as
(04:02):
many deals as we can and hopethat they all go well, hundreds
of investors, hundreds of deals.
We've taken an oppositeapproach, really, where we we
work with deep-pocketed,sophisticated investors,
intentionally global, and limitthe number of them so that we
can really mind meld with them.
And then on the manager side, wetry to find the best managers in
(04:24):
the world, groups like GeneralCatalyst, others that that are
household names, and we buystakes in their top coat.
So we are a minority owner oftheir business, not just an
investor in their funds or aco-investor alongside them,
those, though we do those thingsas well.
And what you get is a kind ofnetwork.
I mean, everybody's talkingthese days.
(04:45):
We're obsessed with the idea oftaking nodes, connecting them
together, and buildingartificial intelligence.
I'm sure we're gonna talk aboutthat today.
I hope we have one of the firstactually interesting
conversations about AI in sometime because it feels like it's
all anyone wants to talk about.
And I hear the same things allthe time.
So we'll try to say somedifferent things.
But one thing that um we dodifferently is we're building a
(05:10):
network, a kind of neuralnetwork, if you will.
But rather than those nodesbeing neurons or chips, they are
organizations and people who areexpert in what they do.
And they're from all over theworld, and they know areas like
energy and you mentionedfashion, right?
We've got a consumer businessthat we back.
We we we look at biotech, welook at fintech, and we work
(05:33):
together not just in the waythat people chat when they're at
Davos, you know, everyone'spuffing out their their peacock
feathers and and trying to bethe biggest organization in
town.
Our strategy is different.
We actually are in businesstogether.
So we're not just meeting at acocktail party.
We own each other's businesses.
(05:55):
That creates a permanentconnective tissue in this sort
of neural network.
And then because the network isbuilt out of people, not chips
or neurons, we can actually gettogether.
And that's when it comes alive.
So every year we host ourcompany, Collective Global.
It's actually hosted by one ofour asset owner, pension, or
sovereign partners.
Last year it was in Stockholm,and we had our major event at
(06:19):
the hall where they give out theNobel Prizes.
This year it'll be in Nice,France, next year in Sydney,
Australia.
And that's when this kind of, Iwon't call it it's not an
artificial intelligence, butthis kind of organizational mind
comes together.
And it helps us all becomebetter thinkers, sharper
thinkers about what's happening,especially in innovation.
SPEAKER_00 (06:40):
So we have to talk about AI.
The three topics that I thinktop of mind for investors,
obviously, AI is one of them.
Robotics, I saw it was describedas the largest addressable
market in human history.
So there's robotics, and thenthere's another kind of code,
which is genetic code.
You're active in all three ofthese areas, but let's start
(07:02):
with artificial intelligence.
We have, you know, we the thediscussion is doomsday or
unlimited promise.
You argue that that's not theway to look at it.
Tell us why.
SPEAKER_01 (07:14):
Yeah, I mean, first of all, I I think if you ask 12
experts on AI what they think,you'll get what's the joke, 13
different opinions.
Right.
And nobody really knows becausethe pace of change is just so
incredibly fast.
I will make a promise here thateven though I'm not a STEM
person per se, I was aphilosophy grad student and then
(07:36):
a lawyer, I think that actuallycan help us cut through some
jargon and talk about uh AI in aslightly different way.
So instead of just throwingaround words like deep learning,
foundational models,reinforcement learning, genetic
token, is it a large languagemodel?
Oh, then one day it's supposedto be what about small language
(07:58):
models and you know, uhmultimodal.
And my favorite word ishallucination.
I mean, what are we really,really talking about here?
So I think that rather thanhaving a simplistic doomsday,
you know, versus utopianRorschach test where you just
say AI and you see whetherpeople's blood pressure goes up
(08:20):
or down, it is possible to havea nuanced view of what AI is
good at.
We are spending roughly 1% ofgross world product, a trillion
plus a year, on the stack ofwhat is AI, from energy through
to data centers, through tochips, through to foundational
(08:41):
models, through to apps, throughto agents, ultimately, in some
cases, as you mentioned, throughto robots and actual
experiments, we're spending aton of money on that.
You would have to go back tomore than 100 years ago.
I think some historical contextis helpful, if you'll forgive
me.
Like when was the last time wedid that as a private sector?
(09:01):
I know we spent 3% of GDP onputting a man on the moon, but
when was the last time that theworld just organically decided
through the minds of investorsand business people to put this
much into one coherent area oftransformation?
I would say, you know, the callit 30 to 50 years where we
invested in railroads,electrification, and
(09:24):
agricultural mechanization,which were obviously three
linked technologies.
The transformation that camefrom spending 1% of gross world
product in those three areasover the call it roughly second
half of the 19th century wasincredible.
Imagine grabbing somebody fromthe year 1800 in America when
(09:47):
90% of people worked on farms,and telling them that in 2026 we
were going to have 2% of peopleworking on farms, and
unemployment would be less than5%.
And then asking them, what'severybody gonna do for a living?
I don't think, I mean, theycertainly wouldn't come up with
management consultant or podcasthost or electrician, right?
(10:09):
Those were not on the menu atthe time.
So I just say that because theamount that we don't know
compared to the amount that wedo know is so vast.
What we don't know is an ocean,what we do know is a drop of
water.
We can say some things aboutthat drop of water, and I'm, you
know, I think we can take outthe microscopes and say some
interesting things about it.
(10:30):
There's a lot that we stilldon't know.
So jargon's not going to help.
It just sells news stories andgets people anxious and maybe
drives up valuations on certaininvestments.
But there are coherent thingsthat we can say about AI.
Certainly we can expecttransformation that is at least
equivalent to the transformationthat we saw coming out of that
(10:53):
tri-pronged era of innovationwith railroads, electrification,
and agri agriculturalmechanization that I talked
about, in which the world becamethe one that we know today.
And the difference here is it'sgoing to happen much faster.
It's already happening muchfaster.
And it's happening in ways thatwill have some predictable
consequences and some second andthird order consequences that
(11:14):
are very difficult to predict.
SPEAKER_00 (11:15):
I saw a brief clip this morning on YouTube.
It was one of the leading AIalgorithm writers, I guess you
would call them.
And he was saying that herecently attended an AI
conference and was himselfastonished by what was projected
to happen, not in the next sixyears, but in the next six
(11:36):
months?
I wonder, as somebody whoinvests in funds that invest in
AI and that, you know, co-investon a number of those deals.
Is the pace of change AI soextraordinary that it sort of
freezes you, if you will, fromfrom investing?
Do you have a pretty good ideaof what you want to invest in
(11:56):
and there's a rationale behindthat?
SPEAKER_01 (11:58):
Aaron Ross Powell It's a little bit of both.
I think you need to be humble inrecognizing that we don't know a
lot, like I just said.
But I also said that, and I madea promise that we'd try to have
a conversation that was coherenthere about AI, which is rare
these days.
And let's start with a theory ofwhat kind of intelligence we're
(12:19):
building.
I think everyone's talking aboutwe're building AI, and they they
feel like that's enough.
Oh, I just said we're buildingartificial intelligence, as if
that's a sufficient description.
It's not a helpful description.
What type of intelligence is it?
What is it good at?
What is it not good at?
When is it going to be good atthe different things that it
could help us with in society?
(12:40):
So let's try to unpack that.
When did we all first hear aboutAI?
Well, I think it's when DeepMind beat Kasparov at chess.
Right.
I forget the year, but that youknow chess was the first
example.
Then AI had a win again in amore complex game, a Go, where
beat the world champ Lee Sadal.
(13:02):
If you put those two data pointson a chart, and now I'm going to
put a third one up there becauseI've just I can report an a very
interesting breakthrough, acompany that I serve on the
board of called Lila Sciences,where they have crafted an mRNA
molecule that in vivo has anorder of magnitude better
(13:23):
results in fighting certaincancers in the human body than
Big Pharma has come up with.
And they did it with AI.
What do those three data pointshave in common?
I think they're examples ofwhere AI is flourishing as a
type of intelligence.
They have in common that theyare all based on rules.
Chess has rules, Go has rules,DNA, mRNA have has rules.
(13:47):
Secondly, you've got immenseamounts of data, just brute
force, more data than a humanbeing can hold in its mind at
any given time.
So the the analogy I like to useis if you've got 10 light
switches on a wall, they can bein a thousand and twenty-four
different arrangements.
If you've got a hundred lightswitches on a wall, there's more
(14:09):
potential arrangements thanthere are atoms in the universe.
Now think of a chess boardrules.
You got huge amounts ofpotential formations of all the
pieces, and you're quickly, youknow, at more than the number of
arrangements than there areatoms in the universe.
Go was supposed to be impossiblefor AI because it was an order
(14:30):
of magnitude or more complexthan chess.
Well, mRNA, of course, exists,and RNA exists because it's how
nature chose to encodeinformation.
And the combinatorial landscape,to get fancy about our language
here for a second, that iscreated through four base pairs
interacting in almost an endlessvariety of ways and then coding
(14:53):
proteins, which can then expressin an almost endless variety of
ways, is shocking.
It's important, and it'ssomething that humans are just
not very good at doing.
We have made great progress.
I don't mean to suggest that thelast 450 years of science
post-Bacon and Descartes andothers in the early 1600s hasn't
(15:16):
been incredible.
But the pace at which we canconquer this particular type of
problem, one with rules, onewith vast amounts of data, and
one with huge combinatoriallandscapes, is is now newly
solvable.
So what's in that space andwhat's not in that space?
Well, I would argue thatcertainly this is going to be
the century of biology, right?
(15:37):
Biology happens to be reallycomplicated.
And so it fits what I justdescribed to a T.
So we've done a lot of investingin enabling AI to access
biological systems.
They often do it through robots,because robots can absorb much
more data.
And biological systems areinherently incredibly complex.
I think there are somenon-biological scientific
(16:00):
systems, like how do we replacethe, what is it, seven rare
earths that China's given ustrouble over in terms of their
use as catalysts in differentindustrial and other military
processes?
Well, that's the sort of thingyou need to run experiments on.
And turns out nature iscomplicated.
You can't just think your waythrough it.
You need to actually go run theexperiments, onboard the data,
(16:23):
analyze the data, synthesize thedata, and then run the next best
experiment after that.
So those are some areas where AIcan be helpful.
I'll give you one more justbecause I don't want both of my
examples to be focused onscience.
You know, AI gets a wrap thesedays.
You talk about the dystopianangle, it's going to replace all
of our jobs, et cetera.
(16:44):
There are also some hugehumanitarian crises that don't
get discussed enough in theworld, in my view, where AI
could, if allowed, be incrediblyhelpful.
The UN is a UN statistic.
There are 154 million orphans.
That is a shocking number.
It's 2% of all humans alive thatare under age 18 and don't have
(17:04):
parents.
Um it's something I knowpersonally because my wife and I
have been very active andadopted a kid.
And uh the reason there are somany is well, it's very
complicated and it's it's it'sgeopolitical and it's it's we
just don't live in a fair world.
But one other reason is thatthere's countless languages,
tons of paperwork, and 154million is a big number.
(17:26):
Well, what's AI good at?
Reading and understanding anylanguage, following any legal or
bureaucratic process, andcapturing and processing
enormous amounts of data.
Think about the amount of moneywe've spent on dating apps to
match, you know, Joe and Sue,who live down the street from
another, one another in, youknow, in Urbana or in Columbus
(17:49):
or in Atlanta.
How about we spent a fraction ofthat matching kids with families
that want kids?
I'm not saying that it's such aneasy problem, but it is a
problem that AI could help withbecause it is a large
combinatorial space that isrule-based with a ton of data.
SPEAKER_00 (18:07):
I wanted to talk a little bit about Lila.
I'm just fascinated.
How did it, presumably you'dnever heard of it like three
years ago, now you're investedin it?
How did how did you come to knowabout it?
What convinced you to invest init, et cetera?
SPEAKER_01 (18:22):
Yeah, I certainly hadn't heard of it.
It was in stealth mode.
But one of our manager partners,General Catalyst, who I
mentioned earlier, and we've gotnow six manager partners.
Collective Global, my firm, isis is relatively new.
I've been doing some variant ofthis kind of investing for 25
years, but we decided to focuson innovation in 2023.
(18:42):
I think our timing was good.
And we backed General Catalystalongside McKinsey, JP Morgan,
and Amazon.
We were one of their backers atthe top co level.
And they've been wonderfulpartners.
And turned out that, if Iunderstand it right, the biggest
safe, which is a particular typeof early stage investment that
GC General Catalyst had everdone in its history was within a
(19:06):
company called Lila Sciences.
So immediately my my ears perkup and they say, You want to see
the future of science?
Go to Boston, check out thislab.
And they send me a picture ofBill Gates, you know,
practically fogging up hisglasses.
He's so excited about what he'slooking at as he's in a lab coat
in the Lila Sciences facility.
(19:26):
The most interesting thing aboutthe photo is that nobody is
touching anything.
So if you think about the wayscience has been done for the
last several hundred years,right, since the age of Louis
Pasteur, right?
We had grad students withpipettes or the equivalent,
moving something from Petri dishA to Petri dish B.
Then they spill their coffee onit.
(19:47):
Then, you know, the their theirthesis advisors argue with them
about it.
Then they tenured professorpublishes something on it.
Then they argue with a differenttenured professor about whether
they were right or this othertenured professor was right, and
did they falsify the data?
And it's it's a mess.
Human science is a mess.
It's also the most productivething we've ever figured out how
(20:10):
to do.
So don't I don't I'm I'm notdemeaning science, right?
I'm just telling you, it's amessy business.
It takes a while.
Now picture instead the insideof one of these semiconductor
slash chip manufacturers, youknow, with the clean hazmat
suits and the 30-foot ceilingsin places like Taiwan.
It's just incredible.
(20:31):
Imagine we were applying thatprecision to the biological
sciences, to the materialsciences in other domains.
Well, that's what Lila does.
And so I could appreciate thatthe entire wheel of science
itself needed an update, andthat putting AI with the human
in the loop at the top of thatchain, and then using robots
(20:53):
that could ultimately synthesizetheir own equipment through 3D
printing to improve recursivelythe quality of the experiments
that they're running, run them24-7, they don't spill coffee,
they don't make a mistake aboutwhich petri dish they just put
the sample in, they suck uporders of magnitude, more data,
they process it faster, andbecause they can hold in their
(21:14):
mind at one time a much largercombinatorial space, like the
space of all mRNA to fight aparticular type of cancer or
other disease, they can progressmuch faster.
And so this is one of thoseareas where whether it takes six
months, as you said, John, or oror or six years or even twenty
years, if you're a big assetowner, that's what I call the
(21:37):
sovereign's pensions, insurancecompanies, and others that that
I work for really, that I helpconnect to innovation, it almost
doesn't matter if it's sixmonths from now or six years
from now.
You've got liabilities to yourpensioners that last 10, 15, 20
years, sometimes longer.
If these disruptions through thesciences or in other areas, we
(21:58):
all witnessed the SAS pocalypserecently.
Which we could talk more aboutlater, where private credit got
thrown under the bus.
If these changes are going tohappen in the life cycle of your
liabilities as a pension orsovereign, you need to be
worried about them.
And it is evident to me with,and I don't have a crystal ball,
but I do have call it 12 to 24months advance warning on what's
(22:20):
happening because of theinvestments that we're making
that are in small privately heldcompanies.
This stuff is coming not justfor new discoveries.
It's going to change the other98% of your portfolio in ways
that aren't always positive.
So you need to be eyes up, chinup, thinking hard about
innovation today in a way thatfive years ago you could maybe
(22:43):
justify not having a venturegrowth portfolio.
Today, I don't think any CIO inthe world can hold their head
high unless they reallyunderstand and make an effort to
understand the pace of change.
SPEAKER_00 (22:55):
Aaron Powell Just as an example, the impact of the
change that is coming on otherparts of one's portfolio.
Are there specific sectors thatyou think are most at risk or
specific companies that youthink are most challenged, I
guess is the polite way to putit?
SPEAKER_01 (23:12):
Aaron Powell Well, yes.
I actually think, for what it'sworth, that this recent SaaS
pocalypse, as I called it,software as a service getting
hammered.
Uh and in particular, I think itwas 40% type drops in the listed
private credit asset managerslike Aries and Blackstone and
Blue Owl was an overreaction,personally.
(23:34):
I think a lot of these SaaSbusinesses are good businesses.
They got hit hard because, asyou know, John Claude Code came
out, and all of a sudden peopleare worried, well, maybe I'm not
going to need to buy Salesforceor some other SaaS business.
And if you're an Aries andyou're lending to them and
you've got a duration on yourprivate loan to that company
(23:55):
that that goes into the 2030s,and all of a sudden you're
thinking that the equity mightbe wiped out, and maybe even
they won't be able to cover alltheir debt in 2032.
Well, that starts to be aproblem, and the market, I
think, overreacted, priced thatin quickly.
My particular view, which mightbe wrong, might be right as to
why that was an overreaction, isthat, you know, let's take
(24:16):
ourselves back in time to whenYouTube came out.
When YouTube came out, itfloundered briefly because
everyone thought, why do we needmore video?
There's tons of high-qualityvideo.
You can stream it this way, youcan rent it that way.
Turns out we had a lot lessvideo than we really needed, as
my tw you know, fifth, sixth,and seventh graders' behavior
patterns every day will testifyto.
(24:39):
So maybe we just have a lot lesscode than we really need.
And so this reaction to say,okay, now companies that provide
code, software as a service, aregoing to die, I think has been
overstated.
So that happens to be one I'mnot worried about, but you are
not as worried about.
You asked me about some areasthat I am worried about.
So here's a big one.
Like what happens whenself-driving cars actually turn
(25:02):
on?
I mean, we we all believe thatthat's gonna happen, right,
John?
I mean, it could be some of it'salready happening, but i i i i
is it gonna be five years, tenyears, fifteen years when this
horrible corridor that you and Ideal with between Fairfield
County and New York City startsto flow like like water instead
of like molasses because it'sself-driving cars.
(25:24):
Well, maybe your publicinfrastructure bets, which you
thought were the safest part ofyour institutional portfolio,
your light rail investments incities like LA or Toronto, which
were built around the car thatyou thought were safe, maybe
they're not so safe.
I'll give you another example.
Your commercial real estateplays in the life sciences,
(25:45):
where you thought, oh, well, ifI if I have if I have wet lab
space near researchuniversities, boy, that's
probably going to be a good 10to 20 year investment.
Well, I've just given you apicture of what science might
look like, will look like in myview, in five, 10, 15 years.
And turns out robots don't likenine-foot ceilings or 10-foot
ceilings.
(26:05):
They need 20-foot ceilings,depending on exactly what's
happening in the space.
So all of a sudden, what seemedlike a safe bet isn't.
And to me, that's the it's boththe returns you can get from
innovation, but also theinsights that you need to have,
the kind of early warningsystem, if you will, that's
gonna predict the seismicactivity in other parts of your
(26:27):
portfolio.
I don't see people talking aboutthat.
Just like I don't see peopletalking about the type of
intelligence that AI is actuallysmart at.
I mean, there's tons of thingsAI is not good at, by the way.
Like I wouldn't pick it to be mysports agent if I were a good
athlete, right?
It's not gonna negotiate a dealand understand all the moving
(26:48):
parts in a complex humansituation.
But it's very good at thesekinds of combinatorial space
thinking that I talked about.
Similarly, there are industriesthat won't be impacted.
I think live entertainment isfairly AI proof.
My kid will still be able toroot for the Yankees in 20
years, even if he won't be ableto play for them.
But there are lots of areas thatwill be impacted.
SPEAKER_00 (27:08):
Aaron Ross Powell There's a group out in
California called Rethink X, andthey did a paper on
transportation as a service.
And so in this model that theybuilt, the cars are electric,
they're self-driving.
You might own one, but you onlyuse it an hour a day, so you
could rent it out three or fourhours a day.
The electric car has 128 movingparts.
(27:32):
Uh combustion engine has, youknow, a thousand moving parts.
So maintenance of that car is,let's put it this way, much less
than it would be for acombustion engine car.
Because it's self-driving andpresumably safer than current
situation, insurance costs willgo low.
That kind of disruption ifyou're looking at
(27:52):
transportation, is thatsomething that that you're
thinking, okay, this is reallygoing to be totally different,
and how do we invest in thisspace?
Or I guess you know you'd justsay transportation's too c you
know too too outside of ourspace, so we'll leave it be.
SPEAKER_01 (28:10):
Well, we we may not be, uh part from our investments
in self-driving cars, which wehave, we may not be focused on
big public infrastructure typeinvestments like the light rail
or or or commuter rail that I Italked about a moment ago.
But our pensions and sovereignsare, and those are their biggest
bets sometimes.
Multi, multi-billion dollarinfrastructure plays that they
(28:31):
thought of, oh, well, this willgive me a five, six, seven
percent yield consistently forthe next 25 years.
Losing steam is a much biggerdeal than missing out on the
next trillion dollar startup forthem.
unknown (28:42):
Right.
SPEAKER_01 (28:43):
Right?
Because they're gonna own asmall percentage of that
startup, but they already own abig percentage of those
infrastructure plays.
And I think what we need to getcomfortable with, and we started
the conversation by talkingabout the sheer pace of change.
Talk about electric cars, right?
You you press the pedal to themetal, so to speak, in a you
know, an electric car and thetorque throws your head against
(29:03):
the headrest.
That's the pace of technologicalchange today.
And outside of wartime in humanhistory, we're just not used to
thinking about the future beingso dramatically different than
the present in a fairly shortperiod of time, five to ten
years, maybe less.
So we've got to be on the ballsof our feet, the tips of our
(29:26):
toes as investors.
And even that will just give usan edge, not uh not a right to
win.
So, but but an edge is is betterthan than sticking your head in
the sand in this moment.
So I'm hoping that more CIOswill pay attention to the
innovation economy and thinkabout these second and third
order consequences for what theythought of as the quote unquote
(29:49):
safe part of their book.
You know, the the deeperquestions here that that you
know maybe maybe we end withare, you know, how is this gonna
even change the role of a CIO,of a decision maker as an
investor?
You know, it's in in every erawe humans have thought of
ourselves as the protagonists.
(30:10):
We're the ones driving history.
And we're now entering an era inwhich that is, I would say at
best uncertain.
And to me, again, you know, youcan look at that and and just
get depressed, or you can lookat that and get really excited.
And probably you can look at itand feel both ways at the same
time.
But I'll tell you what makes meexcited about it is that you
(30:30):
know, we don't really have a atradition that helps us think
about what humans can offer ifit's not being the protagonist,
if it's not being the smartest,the toolmaker, the species above
the others.
And and if we are no longer thesmartest, if we are no longer
the toolmakers, if we are nolonger the species above the
others, because we we gave birthto it in the form of AI, it's
(30:54):
going to force us to ask allthese really interesting
questions about what does, ifanything, make us special?
And I think I get excited about,you know, we started this
conversation talking about my myfamily, and you know, we've
grown it through adoption.
We I've I've also I think thethe greatest single act of
goodwill I've ever done is thatI'm originally from Atlanta and
(31:15):
I've allowed my 12-year-old sonto be a Yankees fan.
Just think about that for aminute.
Right.
What that's real personalsacrifice that involves.
But so so like we've got toreimagine what it means to be a
human in the context of notbeing the smartest, of not
necessarily being the mostefficient, most productive
protagonist uh on earth.
(31:37):
And I think that's gonna leadto, I hope, a revival of of
certain types of arts andculture that I know we both care
deeply about.
And uh I'm excited.
I'm excited both for theeconomic side of things and for
this non-economic side.
SPEAKER_00 (31:50):
Aaron Powell The question people always ask
people like you is you're you'reobviously plugged into a network
of extraordinarily intelligentpeople, but you also read a lot.
What do you read?
SPEAKER_01 (32:02):
As you know, John, we trade books.
So I'm working my way down thebook list you sent me.
But ever since I discoveredaudiobooks, and I can read them,
read them as I'm falling asleep,I read them while I wake up in
the morning, I read them whenI'm on the train, I read them
when I'm between meetings.
I now get a chance to read morethan I have since college.
I felt really guilty about thatfor a while.
I thought it was somehowcheating.
(32:23):
And then I remembered that,well, for most of the call it
30,000 years that human beingshave been telling stories, give
or take, it's been oral, notwritten.
So I thought, okay, I could Icould Homer was mostly, you
know, an oral tradition.
Ultimately somebody wrote itdown, kept it in a in a medieval
monastery so that it could bepassed down to us.
(32:45):
But so once I got over thatshame of being an audiobook guy,
I read pretty much everything,John, and more nonfiction.
I I feel like I've I've come tothe end of what I can read in
physics and uh life sciencesbecause I'm not smart enough to
read the scientific papers, andall of the popular literature
has just become repetitive forme at this point.
(33:06):
So I read a lot of history and Ijust try to find stuff I've
never heard about.
So I just finished a wonderfulbook called Sea Peoples about
the history of Polynesia, andthen I went there with my
family.
It was so great.
So I'm just looking for areas Idon't know about, and then I'll
click on that.
SPEAKER_00 (33:23):
Aaron Powell We've reached, uh I think uh we've
taken enough of your time here,but I wanted to ask you one last
question, which is you'reinvested in AI, you're invested
in companies that do code acrossartificial intelligence,
robotics, genomics, but you'realso invested in fashion.
So the question is why?
SPEAKER_01 (33:43):
Aaron Powell You know, it's funny.
The investment in my portfoliothat my kids are the most
excited about is not LilaSciences, all respect to them.
It's not vast data, it's not anyof these big, you know, unicorns
in the AI space.
It's the fact that we ownindirectly a bit of Justin
Bieber's fashion company.
And I asked my 12-year-old, aspart of my diligence for the
investment, John, I said, Whatdo you think?
(34:06):
Justin Bieber, fashion.
He said, Dad, Biebes is back.
Beeps is back.
Beebs is back.
And so I thought, okay, I'lltake that into consideration.
And then he asked me a question,which I thought was, if I can
brag on my own son for a minute,I feel like that's allowed.
He asked me a question that Ithought, okay, this kid could
have a future in venturecapital.
(34:26):
So he said to me, Dad, does heown the company?
Or is it like one of thoseJordan Nike things where he just
owns the brand?
Well, great question.
He owns the company.
And so then my son said, Youbetter, you better invest.
Piebes is back.
He owns the company.
Go for it.
So we did.
So anyway, I mean, it just tocome to your question,
innovation, it's an it's an easytime to be reductivist about
(34:49):
innovation and just think allinnovation is AI.
It's not.
Like we're also putting nuclearinto space and we're investing
in Justin Pieper's fashioncompany, right?
Like it's it's the one thing AIis actually not doing is
originating truly new content atthe frontier of any field yet.
(35:09):
And that includes fashion.
And so if we can be there withthe best partners in the world
who actually understand thesethings that I certainly don't,
because I started this by sayingI'm not the best in the world at
anything, but I try to be thebest partner.
If we can be there with them,sourcing and then executing on,
and then hopefully adding valueto the best investments in
(35:30):
innovation, it'll span the wholewaterfront from AI to fashion.
SPEAKER_00 (35:35):
I think we've found the uh headline for this podcast
The Beebs is back.
Uh Dan, thank you very much foryour time today.
Uh we look forward to speakingto you again in the future.
Pleasure, John.
Thanks so much.
Popular Podcasts
Betrayal Weekly
Betrayal Weekly is back for a new season. Every Thursday, Betrayal Weekly shares first-hand accounts of broken trust, shocking deceptions, and the trail of destruction they leave behind. Hosted by Andrea Gunning, this weekly ongoing series digs into real-life stories of betrayal and the aftermath. From stories of double lives to dark discoveries, these are cautionary tales and accounts of resilience against all odds. From the producers of the critically acclaimed Betrayal series, Betrayal Weekly drops new episodes every Thursday. If you would like to share your story, you can reach out to the Betrayal Team by emailing them at betrayalpod@gmail.com and follow us on Instagram at @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
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