195 | The AI business application battle is intensifying, Self improving AI, AI for early cancer detection, and more AI news you should know for the week ending on June 6, 2024 - Leveraging AI

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Hello and welcome to a WeekendNews episode of the Leveraging

(00:03):
AI Podcast, a podcast thatshares practical, ethical ways
to improve efficiency, grow yourbusiness, and advance your
career.
This is Isar Metis, your host,and like most weeks, we have a
jam packed episode for you tocover.
We have lots of stuff thatdidn't make it to this week's
news, but you can find all of itin our newsletter, so make sure
to check it out.
There's a link in the show notesto get that.
We are going to cover today fourmain topics.

(00:25):
One is about the new glowing hotbattle ground of ai, which is
applications for enterprise.
Followed by a fascinating reportby Mary Meer, the Queen of the
Internet, about the speed inwhich AI is happening compared
to the internet revolution.
We're going to continue talkinglike in many previous weeks
about layoffs across the boardand AI impact on that, and we're

(00:48):
gonna finalize the deep divesections with talking about
self-improving ai.
And then as always, there's avery long list of rapid fire
items, including somefascinating and really important
scientific discoveries andmethods with ai, like the
ability to better detect cancer.
So lots to talk about.
So let's get started.

(01:11):
In previous shows, we talked alot about the fierce battle that
is happening right now on theturf of writing code with ai,
with all the big playersfighting to grab market share
between companies who build theapplications like Cursor and
Windsurf.
More about them later, and OpenOpenAI, Chachi, pt, Claude, and
platforms like ChatGPT, Claude,and Deep Seek fighting to show

(01:34):
who has the best codingcapabilities.
Well, there's a new battlegroundthat is hitting up very, very
fast, which is the battlegroundof corporate and business
applications in general.
It's not new, but it's justheating up dramatically.
And this week was a greatexample of that.
So OpenAI just announced severalnew different capabilities,
which are all focusing on thatexact topic.

(01:56):
So Chachi PT now startedoffering meeting recording and
transcription, which issomething that was saved to
companies like Zoom and Teamsand Fathom and many others that
are focused on that particularfield.
They also announced newconnectors, connecting straight
into chat GPT and custom GPTsfor Dropbox Box, SharePoint,
OneDrive, Google Drive, and theability to connect MCP servers

(02:20):
to the chat GPT universe aswell.
Those of you who don't know whatCPS are, first of all, check the
episode re-released on Tuesdayof last week.
We did a deep dive into what areCPS and how to connect to them
and how you can benefit fromtheir, in your AI universe.
But in general, they allow youto connect very, very quickly to
data repositories and tools thatyou have in your company, such
as your C-R-M-E-R-P, et cetera,as long as they have an MCP

(02:42):
server and bring them withinminutes into your AI universe,
whether chats or agents, and soOpenAI is focusing very
aggressively on buildingapplications for businesses.
So it's not just that they're afrontier model company, they are
a enterprise tool provider thatis competing with a very wide

(03:04):
range of companies and toolsthat are out there in the
market.
And combine that with the factthey currently have grown by 50%
in just four months to somethingbetween 700 and 800 active
million users.
And they have combined that withthe fact that OpenAI now serves
over 3 million enterprisecustomers.

(03:25):
Grown up from just 2 million inFebruary.
That a 50% growth in just fourmonths of enterprise customers
shows you very clearly wheretheir focus is at right now.
And it's a very serious risk toa huge range of applications,
companies and startups that havefocused on building these kind
of applications on top of chatGPT in the past year and a half

(03:48):
or two years.
Now the only question thatobviously is in your mind and
everybody's mind is, okay, whereis my data going and what's the
control over this data?
OpenAI is claiming and confirmedto, several different news
outlets that they will followthe organization's access
control hierarchy.
Meaning if you are going to usechat GPT in order to connect to
company data, different peoplewill have access through Chat g

(04:09):
PT, to different levels of data,depending on their access
control of their organization.
So if some people cannot see thesalaries of all the employees as
an example, they will not beable to see that through chat g
pt, even though all the datawill be connected.
Exactly how they're doing that.
Is it bulletproof or not?
I think time will tell.
I think we'll know very quicklybecause I'm sure a lot of people
start implementing this.

(04:31):
I can tell you withconversations with several
different people who have triedthis in organizations, they're
saying that it actually workswell and they're only seeing the
stuff that they're supposed tobe seeing.
again, I don't know howbulletproof this is, but in
probably a few weeks we'll knowwhether this is working or not,
and I'm sure they will patch upwhatever is not working right
now because that's obviously acritical component to any
company.

(04:53):
Now, in addition to killing manysmaller startups who build
applications to do exactly thesethings.
They are moving straight intothe fields of giants, right?
These are exactly the thingsthat Microsoft and Google have
promised us to deliver.
The connectivity of all theinformation straight into a
conversation that we can ask anyquestion about any project,
about any process, about anytool, about anything that we

(05:13):
want, and have the data beingcollected from our documents,
from our ERP, from our CRM, fromour emails and so on.
And it seems to be that OpenAIare on the frontier of that,
potentially even ahead ofMicrosoft with their own
internal tools and Google withtheir own internal tools.
Google is obviously not stayingbehind.
Google introduced a lot of newcapabilities in their event a

(05:35):
couple of weeks ago, and they'renow announcing that Google Drive
is launching a catch me Upfeature that allows users to
click the button and get asummary done by Gemini of what?
Change in specific files.
In specific drives.
So if you are using Google'scollaborative environment where
multiple people can access thesame drive and make changes to a

(05:56):
document, and a lot of people, alot of companies are doing that,
then you can go into a specificfolder, click on the catch me up
button that appears on the topsection of the Gemini Bar on the
right side of your screen, andit will tell you exactly what
changed in which document.
I find this feature to be veryuseful.
I don't have access to it yet,but they're rolling it out and
I've seen examples of peopleactually using it, and it looks

(06:18):
like an awesome feature for, asI mentioned, anybody who's doing
collaboration with Gemini, theyintroduced a lot of other
capabilities, as I mentioned acouple of weeks ago.
And it's supposed to becomeavailable to everybody in the
next few weeks as long as youhave a workspace business
Standard Plus or EnterpriseStandard or Plus or education
licenses of Google, and ofcourse the Google one AI premium

(06:38):
subscription as well.
Microsoft also made a big movewith the release of agentic
retrieval on Azure.
So their new agentic retrievalconcept they're claiming
delivers 40% improvement inanswer, relevant and accuracy
compared to traditional ragsystems.
So the differences is obviouslyinstead of just counting on the
embeddings and the data in avector database like traditional

(07:00):
rag, there's an actual agentsystem and now I'm quoting
Autonomously plans and executesretrieval strategies for complex
questions by breaking down userquestions into focus subqueries
that run in parallel across bothtext and vector embeddings.
Basically, what that mambo jumbomeans is that instead of trying
to do.
The search is it tries tounderstand exactly the

(07:21):
information you're trying tofind, and then it breaks it down
into multiple additionalsearches, which does two things.
One, it allows you to run inparallel, which makes the
process significantly fasterthan doing a traditional rag
search on a big piece of data.
Those of you who tried it knowthat it's really frustrating
because sometimes you wait fiveor seven, five or seven minutes
to get the answers, and I don'thave the same amount of data
that large organizations do.

(07:42):
and the other thing is itactually provides better results
because it understands thecontext and what you're trying
to get, and it does betterqueries across multiple sources.
This is currently in initialbeta testing.
It's not available to everybodyyet, but it's rolling out.
And so anybody with an Azureaccess and data on Azure will be
able to use it in the nearfuture.
But there are other big playersin this field that are

(08:02):
aggressively shifting into AIapplications for enterprises.
Two of the most notable ones areprobably Snowflake and
Databricks.
Both companies or a database,data management, enterprise
level platforms.
That's what they were until AIcame out.
And now both companies are allin on AI solutions for
enterprises on top of the datajust a year ago, snowflake was

(08:23):
not even in the AI conversation.
Databricks got a lot more news,but now they're delivering
already an impressive AIsolutions that has pushed their
earnings by 14% with aprojection of$4.3 billion
revenue guide for 2026, which isshowing you that they're very
confident that they're moving inthe right direction.
Both them and Databricks arepushing hard on what they call

(08:46):
systems of intelligence.
Basically a layer ofintelligence above your data
that allows to connect the dotsin a way that is very hard for
humans to do both in means oftime and in means of efficiency
and understanding what dots toactually connect.
They're calling it the fourdimensional business
intelligence, and the idea is togo from static dashboards that
just shows you the data whichallows you to answer four

(09:08):
critical questions.
Question number one is whathappened?
Question number two, which ismore important why it happened?
Question number three, what willhappen?
And question number four, whataction to take, which basically
means that instead of justhaving data, and you have to
figure it out, you have anin-house 24 7, 365 consultant
that looks at everything thatyou're doing and can provide you

(09:29):
guidance across any vector onwhat actions to take because of
the underlying problems andissues and opportunities in the
data that you have in yourbusiness.
Now, Databricks went an extrastep and acquired tabular, and
that gives them access to a muchwider range of data sets and a
very strong hold into the opensource universe, which now means

(09:51):
they can play both in closedsource and in open source data
management and AR layer on topof that.
And we've discussed similarapproaches from other giants
like AWS with their SageMakerplatforms.
That allows quickly andeffectively creating vertically
integrated platforms that looksinto multiple data points within
your data on AWS Salesforce,very aggressive transitions to

(10:11):
agent force basically switchingtheir business model from
selling seats as a SaaS companyto monetizing the actual
feedback loop and actions thatare happening in companies.
Now, if I go back for a secondto Snowflake and Databricks,
their big summits we're right inthe middle between their two big
summits.
So one of them just had theirsummit this past week.
The other one is having itssummit next week, back to back,

(10:33):
not by chance.
And I'll review all theannouncements from those at the
next week's news.
But what this is telling all ofus is that the battle is now not
about the underlying model,which is still going on.
We're still gonna see open AIand Anthropic and Deep Seek and
Gemini.
All of those come up with newmodels and trying to up each
other.
But the real difference to ourday-to-day life is what we can

(10:55):
actually do with it.
And running our business datathrough it and getting insights
and making better, fasterdecisions is the key to success
in business.
And hence, we are gonna seethese companies go more and more
into areas which were not theircore expertise before, to gain
market share in this new futureway of doing business.
So what we're going to see iswe're gonna see the lines

(11:17):
blurring between a lot ofcompanies who did very distinct
things, where all of them aregonna do a lot more stuff and
gonna go into other companiesareas, trying to capture more
market share in this new way ofdoing business with AI as a
critical layer for making betterbusiness decisions.
And it will be very interestingto see who comes out on top in
this new hierarchy.

(11:37):
Open AI is definitely creepinginto the worlds of giants, but
some of the giants are creepinginto each others fields as well.
And there's obviously a lot ofother smaller startups in that
whole mix.
So it'll be very interesting towatch this field and we'll
obviously keep you updated.
And just another anecdote on howhot this field is, u.com, which
is an AI search company.
We've interviewed their CTO andco-founder Brian McCann back in

(12:01):
episode 1 44.
You should go and check thatout.
What they have shifted veryaggressively from search for the
masses to search for theenterprise, and they are now in
the process of potentiallyraising 700 to$900 million at a
$1.4 billion valuation, which isshowing you how much need and
how much investors believe inthe strength of enterprise level

(12:23):
search and AI data analysis.
Our next topic is the reportfrom Mary Meeker.
So those of you who don't knowMary Meeker, she is the founder
of the VC firm Bond, and she wastagged as the queen of the
internet back in the internetboom days, and she was known for
releasing really large, reallydetailed reports about trends of
the industry.

(12:44):
she hasn't released any reportin the past few years, I think
since 2019, but she justreleased a report called Trends
in Artificial Intelligence.
It is a 340 pages of a report,very, very detailed.
We are obviously not going to gointo all of it because otherwise
it's gonna be three episodestalking just about that.
We're gonna cover some of thekey findings and the key things.

(13:04):
The number one thing that she'shammering is the pace in which
this technology is moving andhow much faster it is than
anything that we've seen before.
So she's talking about the paceand scope.
So the pace and scope of changerelated to AI technology
evolution is unprecedented.
So she's giving a few examples.
Chachi PT reaching a hundredmillions, faster than any other

(13:27):
company.
Chachi PT reaching 800 millionusers in just 17 months.
Nothing even remotely close tothis ever happened before.
The number of companies that arehitting very high annual
recurring revenue numbers isalso unprecedented and never
happened before.
The pace in which competitorsare matching each other's
features and capabilities inthis industry is also
unprecedented.

(13:47):
And she's mentioning that theability of open source universe
to catch up and in some caseseven overtake some of the closed
models in specific things, isalso something that never
happened before.
The only one area that is notoutpacing previous technology
revolutions is returns, right?
These companies are spendingbillions into infrastructure and
into running faster and intotraining models, and so far,

(14:09):
returns are far behind thereturns that we've seen in
previous technologicalrevolutions.
However, everybody believesright now that's gonna come and
come at a much highermultiplier, and hence why
they're pouring all thesebillions of dollars into ai.
A few other key things that youmentioned, first of all, is that
the cost of AI usage plummets.
Basically inference costs for AIusage have dropped 99% over two

(14:32):
years.
So the same level of informationthat you got two years ago,
paying a dollar you're notpaying 1 cent for.
This is very dramatic.
The flip side, training costsare soaring, so we know that the
training costs are going higherand higher and higher.
Current level of models aretrained at around$1 billion for
a training of a new model, whichis showing you how hard it is to

(14:54):
compete in this world where therate of use so people are
actually paying you to use themodel is dramatically shrinking.
While the cost of putting newmodels out there is increasing,
which connects back to ourprevious point that I think
we're gonna see more and moreapplication layer innovation and
integration rather than justcompeting on the underlying
model capabilities, she's alsomentioning that the energy

(15:16):
efficient is improvingdramatically.
She's saying that NvidiaBlackwell's GPUs, their latest
ones from 2024, uses 105,000times less energy per token than
it's generating, than its 2014Kepler GPU from 11 years ago.
Now, while this is verypromising, a hundred, 5,000
times less energy, the problemis we have millions more of

(15:39):
demand.
So the demand has grown inseveral orders of magnitude more
than the energy efficiency theactual underlying infrastructure
is providing.
And so while, yes, this isgreat, I think we still have a
very serious issue with energygeneration to support AI and
with its impact on emissions andglobal warming.

(16:00):
So if you want, this report isfree, you can go and find it,
drop it into notebook like Idid, and you can see the
summary.
You can ask questions, you canget a quick podcast.
It's not that quick, I can tellyou that.
But it's still a very goodoverview of what's happening in
the AI industry right now,especially coming from somebody
who has done similar research onprevious technological
revolutions.
It seems that we can't have aweek go by without talking about

(16:22):
additional layoffs and what'shappening right now.
Well, massive layoffs in the USthis year.
275,000 jobs were cut just inMarch of 2025.
A big part of it, 216,000, soabout three quarters were driven
by Trump's administration,department of Government
efficiency, doge, that was ranby Elon Musk that since then

(16:43):
left that position andcompletely trashing the
administration.
But that's a whole differentthing we're gonna, we're not
going to even dive into.
But 275,000 job loss in onemonth is a lot.
And we talk previously aboutKlarner, CEO, talking about
their 40% reduction in headcountin the past couple of years
because of ai, Shopify, CEO, tobuy slot key, with his memo to

(17:04):
employees saying that theycannot hire new employees unless
they prove that AI cannot do thejob.
Walt Disney just announcedcutoffs in several hundreds of
jobs globally.
Online education form chegg cutto 248 employees, which doesn't
sound a lot, but that's 22% oftheir workforce.
Amazon is eliminating jobs intheir devices and service unit

(17:25):
proctor and Gamble justannounced a job cut of 7,000
jobs, which is 15% of itsnon-manufacturing workforce.
It's gonna happen over the nexttwo years.
Citigroup is planning to reduce3,500 positions, specifically in
China.
And we talked previously aboutMicrosoft cutting 6,000 jobs,
meta cutting, 3,600 jobs.
Workday slashing 1,750employees.

(17:47):
Salesforce reducing a thousandpeople from their head count,
Autodesk with 1350 and so on andso forth.
The list is long as the numbersare very, very big and they're
staggering.
Now, to be fair, this is notjust ai, right?
The layoffs are aligning withglobal economy uncertainty with
the tariffs, war of Trump, withthe war between Russia and the
Ukraine, with uncertainties onwhat's going on in China and

(18:10):
Taiwan and in China's economy.
So there's a lot of otherfactors.
It's not just ai, but AI isdefinitely in the back of that
and it's no longer a secret.
And as I mentioned, many CEOsare saying it out loud that they
are increasing the quoteunquote, efficiency of the
company by making AI take moreand more jobs.
Now, to be fair, there's anarticle from this week that is

(18:31):
showing that the tech sectorlayoffs are actually slowing
down in 2025 compared to 2024.
the midyear number right now isthe 137 companies have cut
62,000 jobs, which means if westay on this space, we're at
145,000 jobs for the end of theyear compared to 152,000 from
2024 and 264,000 for 2023.

(18:53):
So 2023 have seen The most jobcuts then 2024, and now 2025 at
the current pace is actuallyslightly slower.
But the thing is, this is not animprovement because more jobs
are being lost.
It's not that now jobs are beingcreated to offset the jobs that
were lost.
It's just more jobs lost just ata smaller pace, which is
obviously still not good news.

(19:14):
As we shared last week, darioAmadei, the CEO of Anthropic
finally came out and said, yes,this is happening.
AI is gonna take jobs and theleaders of the world need to
address it.
I haven't heard anybodyaddressing it yet, but I think
it'll become more and more of amainstream conversation.
I started having people askingme about it because they've seen
the news about Dario in everynews outlet and asking me what's
going on, and most people don'tstill understand the impact the

(19:38):
impact that AI is going to havebecause they don't understand
what it can do and the fact thatyou're listening to this podcast
and probably other podcaststells me that you're into AI and
you're trying to learn andunderstand, but that's not the
common in the society.
The amount of places I go towhere people just heard of Che
Chi Piti maybe played with it totry to answer an email, and
that's it is probably more thanI see the opposite of people who

(19:58):
are all in like me, who teststuff all the time and
experiment with AI and integrateit across everything that
they're doing.
And so we're still very farbehind on the understanding of
what's the impact is going tobe, and I think it's gonna hit a
lot of people with a very bigsurprise.
Now, to add on top of that, andthat's gonna be our last deep
dive component.
We talked several times in thepast about the point when the AI

(20:19):
acceleration is gonna go throughthe roof, and that's gonna be
the point where AI can startself-improving, basically write
its own code and acceleratevery, very fast.
Where every generation of AI canwrite better code to write the
next better generation of ai.
Where Sakana ai, which is aresearch company in the AI
field, just built what they callDarwin Global Machine or DGM,

(20:42):
and they announced on May 30ththat it now auto autonomously
rewrites its own Python code,and that it was able to boost
its performance by 50% onseveral different benchmarks
between the different versionsthat it's creating on its own.
Now, as the name suggests, withDarwin in the name, the way this
works is it works like theDarwinian evolution concept

(21:03):
where the machine builds severaldifferent variations of the code
and then it tests thesevariations of the code to see if
any of them is better than theexisting code.
If it is better than theexisting code, this becomes the
next version of the AI and allthe other ones get trashed and
so on and so forth, and it justkeeps on going.
Now because it can write codefaster and deployed faster and

(21:23):
check for errors faster and findmistakes faster, it can do this
process way faster than humanscan across multiple aspects of
the code, on one hand, it isreally amazing.
On the other hand, it's reallyscary because the day where this
will be the common thing formost AI platforms is coming now,
it may not come tomorrow.
They're probably doing it on amuch smaller scale then let's

(21:43):
say GPT-4 0.5 or CLA four.
But the concept is there, it'snot being tested and proven.
And that's gonna trickle intoprobably all AI companies.
And that to me is a very scarythought because even today we're
finding it hard to control AIand to keep it in a box and to
verify what it's doing.
We talked about last week, aboutthe deceptive behaviors and more

(22:05):
about that at the end of thisepisode.
But combine that with the factthat it will be able to
self-improve and write its nextvariation very quickly.
This in my little head meanstrouble because we won't be able
to control what it's doing, whatit's improving, because the pace
is gonna be faster than we canactually monitor and update.
And the fact that this is a racebetween companies and the fact
that there's many billions ofdollars involved, it's gonna

(22:26):
drive this regardless of thepotential implications.
and now to the Rapid Fire Newsof the week.
We'll start with Microsoft.
In a UK government trial with20,000 civil employees using
Microsoft 365 copilot isclaiming that they're saving on
average 26 minutes per day perworker.

(22:46):
You put that together, that'sroughly two weeks every year per
every employee.
Now, they've used copilot todraft documents, manage emails,
schedule meetings, creatingpresentations, and streamlining
other routine administrativeactivities across 12 different
government organizations.
82% of the employees whoparticipated in this research

(23:07):
reported that they would notwant to abandon the AI tools
after the research is overindicating that most people were
very happy with using the toolsand the way it helped them be
more productive.
Now, the big question that I'masking every time I'm seeing one
of these surveys, and if youremember, I shared research with
you about this topic a coupleweeks ago, is what was this time
used for?
Meaning those 26 minutes peremployee, what were the gains

(23:30):
and benefits?
The fact that you saved time isawesome, but how was that time
used to make the organizationmore effective, more productive,
and so on is not mentioned inthis research, as I told you, in
the research that I shared withyou two weeks ago, there is very
strong evidence that in thecurrent structure of things,
this save time goes to wastebecause it's not aggregated into
meaningful time.

(23:50):
If you're saving two minuteshere, five minutes there, 20
minutes here, seven minutesthere, it's not a time that
you're gonna put together intostarting a new task.
It might be a time that you goto the bathroom, you go to grab
another coffee, you have a chatwith your coworkers, which are
all important things, but theydon't necessarily drive
efficiency for the organization.
So what I think will happen isthat we'll have to figure out
new ways and new processes, howto work with AI so that time is

(24:13):
more effectively aggregated sowe can actually benefit from it.
Staying on Microsoft.
If you are a video creator,you're gonna love this piece of
news.
So Open AI's, SOA is nowavailable on Bing for free.
Now it's limited, you can onlygenerate five second videos.
And currently it's only nine by16 vertical videos for TikTok,
YouTube shorts, et cetera, withsupport for 16 by nine.

(24:34):
So the portrait, so thatlandscape format is coming soon.
As a free user of Bing, you'regetting 10 fast generations for
free, and then additional videoscost you a hundred Microsoft
Rewards points or will take youhours to render because you will
stay in the queue.
So there's an incentive toengage in the Microsoft
ecosystem to get the rewardpoints, to be able to do more

(24:55):
videos fast.
Now, these videos carry the Ctwo PA digital watermark, that
identifies them as AI generated,which I think is a great step in
the right direction.
The problem is right now thereare several different standards
and not everybody's followingthem.
So it's still very confusing andstill not very helpful to know
what's real and what's not.
Now, this obviously doesn'thappen in a vacuum.
Google released VO three acouple of weeks ago.

(25:16):
VO three is mind blowing and outof this world.
And if you haven't seen VO threevideos, go and check anything on
the internet.
Just search for VO three videoson Google or on any other thing,
like an AI tool.
And you'll see hundreds ofincredible videos that will blow
your mind.
And so Microsoft didn't wannastay behind and they're
releasing Sora for free.
So you can generate AI videos onthe Microsoft universe as well.
But while they're moving forwardand releasing a lot of amazing

(25:38):
stuff with ai, and if you don'tknow what I'm talking about, go
back to the news episode twoweeks ago after their big event.
Not everything is going great.
And while Microsoft is makingvery big announcement and
releasing very impressive AIcapabilities, there's still a
lot of termil internally inMicrosoft and they're making
another restructuring on who'sgonna manage what.

(25:58):
This is the third major reorg ofthe AI Microsoft team since the
beginning of 2024.
If you remember last MarchDeepMind, if you remember last
March, Microsoft acquiredinflection and its co-founder
and the team inflection.
if you remember last March.
Microsoft acquired inflectionco-founder, Mustafa Suleman and

(26:19):
most of the inflection team tocreate a new Microsoft AI
organization.
Last October, they hired formerMeta Head of Engineering J
Paric, to be the AI apps sar,and now there's a whole new
change going on as well.
So the Chief MicrosoftCorporation for LinkedIn is
taking charge of the team thatwill build email and
productivity apps.
Ryan Lansky, the professionalnetworking site Chief Executive

(26:42):
Officer since 2020 will tack onresponsibility for the teams
behind Outlook, word, Excel, andthe rest of the office bundle.
He'll report to Rajesh Ja, whois a top engineering executive,
who is organization includesWindows and business software
and Lansky, as the CEO ofLinkedIn will still report
directly to Satya Nadela.
So he is gonna be wearing twohats, reporting to two different

(27:03):
people.
Now the dynamic 365.
Corporate vice President CharlesLamana will also join JA
organization as well.
So lots of big shifts in verysenior leadership and the
reporting lines that all aretied to how are things developed
within Microsoft to connect alltheir office suite and other
things that they're doing to ai.
That's never a good sign whenyou do a lot of reshuffles,

(27:26):
especially with these newreshuffles are not mentioning
anything about Mustafa Sulemanand his role in this whole
thing.
Exactly.
What does he oversee right now?
Is it just the development ofnew models for Microsoft?
Unclear.
But I will update as the dustsettles.
So on one hand, Microsoft isdelivering amazing new
capabilities and vision and onthe other hand not very clear

(27:47):
what's happening thereinternally right now.
And from Microsoft tophilanthropic.
We'll start with some not verygood news for philanthropic.
So Reddit just filed a lawsuitagainst the philanthropic for
breaching contract and usinguser data without consent.
The lawsuit claims thatphilanthropic attempted to
access the platform's data morethan a hundred thousand times
between July, 2024 and May,2025.

(28:10):
Despite the digital restrictionsand the formal requests that
were sent to Anthropic to stopthis behavior.
Now, we had many lawsuitsbefore, mostly from different
publishers against the big modellabs.
But this is the first time thata big tech company he's suing
one of the giants for trainingon their data, or at least
attempting to train on theirdata.

(28:30):
Where will this go?
I don't know, but this is justanother angle where we see this
battle between the people whocreate and own the content to
the people who believes thatthey can use the content as they
wish.
If you want the exact quote fromReddit, they said, despite what
its marketing material says,anthropic does not care about
Reddit's rules or users.
It believes it is entitled totake whatever content it wants

(28:51):
and use that content however itdesires with impunity.
Now if you want to make thestory a little more interesting,
Sam Altman, the CEO of OpenAI,owns 8.7% of Reddit making him
the third largest shareholder,and he was once one of the board
members of Reddit as well.
OpenAI themselves have alicensing deal with Reddit, and

(29:13):
so this might be a play byReddit.
Now, I'm not saying Sam isinvolved in this, but it's a
very obvious play by Reddit toget another licensing deal.
As I mentioned, they alreadyhave one with OpenAI and they
have a similar agreement withGoogle, so I believe they'll be
able to twist the arm ofAnthropic to do the same thing
if they want access to Reddit'sinformation.
Now, the response from Anthropicwas very generic.

(29:34):
We disagree with Reddit's claimsand will defend ourselves
vigorously.
What does that mean?
I don't know, but as thisevolves, I will let you know.
Staying on Anthropic.
They just announced on June 5ththat they're launching claude
gov, which is a specialized AImodel for US defense and
intelligence agencies.
It is similar to, and competingwith a system that was released

(29:55):
by open AI called Chachi pt gov.
In their statement,philanthropic revealed that
Claude gov has already beendeployed for several different
agencies at the highest level ofUS National Security.
Though they didn't specify whichagencies it was deployed and
exactly when, and unlikeobviously the consumer version
that we have access to, it has alot less guardrails and it's
running within a securedenvironment that allows the

(30:18):
government to use it with hugeamounts of data without the risk
of that data falling into thewrong hands, connecting some of
the dots for you.
On November, 2024, theyannounced a collab on November,
2024, philanthropic announced acollaboration with Palantir, who
has been providing multiple AIplatforms to the government
before on AWS Secure Cloud.
So it's a tight partnershipbetween several different

(30:40):
companies who are now providingsecure AI capabilities for the
government.
Now, where does that lie withthe company's core values when
it comes to philanthropic?
they're saying it's perfectlyaligned with it because their
current guidelines are sayingthat legally authorized usage
for things like foreignintelligence analysis is allowed
while it prohibitsdisinformation, weapons

(31:03):
development, censorship, andmalicious cyber operations.
In my eyes, this is a veryslippery slope, especially once
you're giving the governmentaccess to a secure environment
that you may not have access to.
Even though it's yourtechnology, you don't really
know what they're going to useit for.
And so this is a very, very bigproblem, and I shared with you
my thoughts before, just likethe race on the civilian side,

(31:26):
the race on the military side isgonna be as fierce because
nobody will want to be the onethat doesn't have this
technology and allow the otherside to have it.
So having ai, autonomous weaponsystems and so on is something
that's coming.
It's coming fast and it's gonnachange future wars dramatically
with AI making decisions insteadof humans making decisions.
and that opens a whole kind ofworms that you can think about

(31:48):
on your own.
Or maybe we'll do a wholeepisode about that.
But it's definitely notsomething that I'm happy about.
Staying on Anthropic Anthropicjust launched Anthropic
Explains, which is a blogprimary written by Claude.
There is human oversight that islooking at the outputs and
making sure that they're alignedand making final changes, and
the humans are in charge of thefinal product.
But most of the blog is writtenby Claude itself, and it

(32:11):
includes articles on a widerange of things, including
highly complex topics such assimplified complex code bases
with Claude is one of thearticles that Claude itself has
written.
That comes to show you severaldifferent things.
One, how good these tools arebecoming, and I actually really
like the way Claude writes blogposts and longer pieces of
content and short pieces ofcontent as well.
To be fair, it also shows youthe importance of this

(32:33):
collaboration between humanwriters and oversight to AI
writing the initial content, andthat's the way philanthropic is
following, at least right now,or at least this is what they're
saying.
I can tell you that as thesesystems get better and better, I
think humans will trust themmore and more.
We'll verify less and less, andwe're gonna get information and
assume it is correct andaccurate and is aligned with our
needs and values because we justare going to stop reviewing what

(32:57):
it's doing.
I don't encourage that.
I just think that's humannature.
But to show you the power ofthat and the level of content
that it is generating.
Claude's posts are gathering200,000 plus unique reads in its
first week.
I don't know many blog writersthat ever got that.
Now obviously it's anthropic,they have a big following.
Obviously there's the wholething of, I wanna read the blog

(33:17):
post that was written by Claudeto see how good it is and see if
I can tell the difference.
But it's still a very largenumber of people who have read
the blog that was written, notby a human.
Staying on Anthropic and itsrelationships with other
companies.
Anthropic just cut Windsurfdirect access to its models.
So we talked about Windsurfseveral times in the past.
They are one of the leadingprofessional vibe coding

(33:38):
platforms out there.
There are serious rumors thatthey're gonna get acquired by
OpenAI for$3 billion.
And so because of these newsthat apparently are
materializing philanthropic,decided to cut their direct
access to all their models.
Windsurf, CEO went to X andbasically complained about the
whole situation, said that itdoes not understand this move.

(33:58):
Said that they're willing to paymore to get access to the models
and that the fact that they cutthem off within only a five day
notice is completelyunprofessional and unfair.
And on the other hand,philanthropics co-founder Jared
Kaplan told TechCrunch.
I think it would be odd for usto be selling Claude to open ai,
which makes perfect sense to me.
And Philanthropics, Steve Minchsaid, we're prioritizing

(34:19):
capacity for sustainablepartnership that allow us to
effectively serve the broaderdevelopment community.
Not to be fair, you can stilluse Claude's models on Windsurf,
just not directly from Windsurf.
You have to bring your own APIKeys, which is not a big deal.
The problem is it's a lot moreexpensive than running it
through the Windsurfenvironment, and I'm sure
they're going to lose clients toother similar platforms out

(34:42):
there, such as Cursor.
From Anthropic, let's switch toopen ai, which is potentially
maybe acquiring Windsurf.
And this topic is just a quickmention.
A new book was just released byinvestigative journalist Karen
Howe, and the book is labeledEmpire of ai, which is following
everything that happened fromOpen AI's launch to becoming a

(35:02):
$300 billion fraud profit giant.
So how began covering OpenAI in2019 for MIT technology review?
She has interviewed multiplepeople.
She had access to a huge amountof documents and she has a very
unique view into everything thathappened and happens in OpenAI,
which is obviously a fascinatingread.
I didn't get a chance to read ityet, but I'm letting you know

(35:24):
that it exists and if you wantto get a lot more insight
information on what has evolvedand what has happened in OpenAI
from being a small research labnonprofit for humanity to be
becoming one of the mostpowerful companies in the world,
then I think it's a must read.
And now two specific news fromOpenAI.
OpenAI just opened its Codexcoding agent to Chachi Pity plus

(35:46):
users.
So if you're paying 20 bucks amonth, you now have access to
that.
Before that, it was onlyavailable in enterprise teams
and pro tiers.
So now it is also available tomost of the people who are
paying for Chachi pt.
Now it's available to all thepeople who are paying for Chachi
PT Codex can now connect to theinternet, which is also a new
thing, which allow it to findpieces of code and dependencies

(36:07):
and instructions and APIdocumentation and so on, on its
own.
It also allows it to connect tostaging servers and running
tests with external resourcesand so on.
So it's becoming more and morepowerful.
Going back to what I said in thebeginning, this is starting to
look more like an IDE andcompete with tools like Cursor
and Windsurf, and definitelyit's built in the first step in

(36:27):
order to compete with the codingcapabilities within Claude,
which are very impressive.
I'm actually, I've actuallybuilt several different things
in Claude already.
Mostly simple stuff, dashboardsand games with my kids.
But the coding and executioncapability in Claude is really
amazing.
And so this it's not surprisingthat OpenAI is allowing now
access to everybody to useCodex.
I haven't tried Codex yet, but Iam going to compare the two and

(36:49):
give you my opinion sometime inthe next few weeks.
OpenAI is also planning tolaunch all three pro, which
similar to the all three thatwe're all using only, it's gonna
be only offered to the prosubscribers who pay$200 a month
and it'll provide additionalcomputing power, but also
enhanced reasoning capabilitiesbeyond the basic model.

(37:10):
Well, is that gonna be the thingthat's gonna drive more people
to pay$200 a month?
I don't know.
I actually really like O three.
I think it's very powerful andit's doing a lot of things, but
I don't think I will pay 10 xjust to get a little more of
that.
And some good news also for thefree tier users of OpenAI.
OpenAI is just rolling alightweight memory feature for
the free tier.
So it works very similar to theway memory works for the paying

(37:34):
users, but it only has shortterm memory instead of long-term
memory.
Meaning it's gonna bring thecontext from your recent
conversations into your newconversation, but it's gonna
forget that information that wasolder.
I actually find the memorycapability to be very helpful in
CIP pity, and I hope I can haveit in some of the other
platforms in the same efficiencythat Chachi Pity does it.

(37:54):
So this is really good news forthe free users.
This is not available to peoplein the eu, uk, Switzerland,
Norway, Iceland, and LichtenLichtenstein due to strict AI
regulations in those places.
And if you are a free user, youcan disable the memory and
control what it's actuallyremembering, just like everybody
else.
And from open AI to Apple, whichwe'll probably talk a lot more

(38:15):
about next week because theirworldwide developer conference
is happening this coming week.
But they just shared that theyare testing a large language
model with 150 billionparameters, which per them and
I'm quoting, approaching thequality of recent chate
rollouts.
Now Apple is actively testingdifferent levels of this model
with 3 billion, 7 billion, 33billion and 150 billion

(38:38):
parameters using an internaltool they call playground that
allows them to benchmark theirmodels against other models,
mostly Chachi pt.
Now the large model, the$150billion model runs on cloud
computing similar to other largelanguage models and is obviously
outperforming the on-device 3billion parameter of Apple
intelligence.

(38:58):
We talked about this a lot onthis podcast.
Apple's performance on the AIfield has been nothing short of
embarrassing so far.
They haven't released anythingsignificant.
The things they have releasedare very small components of the
bigger picture that they'vepromised.
A lot of people have bought newApple devices and have upgraded
their iOS platforms to getcapabilities that did not show
up.

(39:19):
There've been a termil inleadership over there.
People removed, new people werehired and they're basically
scrambling to figure things out.
The new series that was supposedto be released last year will
maybe be released in 2026 and itmight get pushed to 2027.
this is not looking good forApple.
And a part of theseannouncements, Google Gemini is
expected to join Chachi PITI asa Siri backend alternative for

(39:43):
iOS 26 with talks also going onwith perplexity to potentially
be a part of Siri and maybeSafari search.
Well, as users it means you'regonna have more options, which
is a good thing.
It is also very clear to me thatthe new partnership between
OpenAI and Johnny Ive is a bigthreat to Apple with its iPhone
dominance in the market.

(40:04):
And I think they're terrified ofthat.
So I think they're looking forpartnerships that are not with
OpenAI to have some otherdifferentiators to stay
relevant.
And there's also two trialsgoing on.
One is against Google Monopoly,which part of that is trying to
break their$20 billion a yeardeal with Apple to be the
default search on Apple'sdevices.

(40:24):
So that might be another reasonfor Apple to move away from
Google search on their devicesto other solutions.
And as I mentioned, theirlargest conference of the year.
The WW DC 2025 is starting onJune 9th, and it will be very
interesting to see how much AIthey're actually gonna be there.
I actually believe there's gonnabe a lot less than we've seen
last year because of the veryserious backlash after last year

(40:47):
was everything Appleintelligence, and they failed to
deliver on everything theypromised.
So I think it's gonna be a lotmore low key when it comes to AI
this year, but I will update younext week.
One additional anecdote onApple, they just drop from third
to fourth place on Fortune 500Company with UnitedHealth Group
overtaking them, followingWalmart, Amazon following

(41:08):
Walmart and Amazon, which arestill holding the first two
places.
And I heard a very interestingpodcast this week that is asking
if Apple is the next Nokia.
Now I know that sounds insane.
They're one of the most lovedcompanies in the world.
They have the iPhone and theiPad and the earbuds and the max
and so on.
And yet Nokia was in the sameexact situation before the

(41:30):
collapse.
They were the largest behemothin the mobile cellular world by
a very big spread.
And I don't know anybody who hasa Nokia phone right now.
Why?
Because they missed the trend ofsmartphones and they were not
fast enough to adapt.
And when they did adapt, it wastoo little and too late.
Now, as the same fate is gonnahappen to Apple, I don't know,
but it's definitely not lookinggood for Apple in the last

(41:52):
couple of years.
Combine that with the fact that,as we mentioned, Johnny, ive now
has a partnership with OpenAI tobuild something that may take
away market share from phones.
Combine that with the fact thatmeta has actually built glasses
that people actually use andlove versus the really
sophisticated, really advancedheadset that Apple developed
that very few people boughtbecause it was three thousand

(42:13):
dollars.
And really heavy combine thatwith the fact that there's a
trial going on against open AIwhen it comes to allowing people
to open a secondary up storethat Apple doesn't control and
doesn't it cannot take 30% ofthe profits off the top.
And that they're facingpressures from multiple
directions and, facing pressuresfrom multiple directions, and
this may lead to results thatare very different than what we

(42:33):
know from Apple right Now.
That being said, their stock isseem to be holding pretty steady
despite all these issues.
But if you look at Apple in thepast decade, there's been very
little innovation again, theonly big thing that you can talk
about that was like, oh my God,this is a new thing, was the
headset, which was a hugefailure.
And so no big innovations comingfrom Apple.
The company who maybe hadinnovation as its main driver,

(42:56):
introducing new things to theworld that didn't exist before.
And now one of the people thathelped them lead that is.
Running against them and they'refacing a lot of other issues.
So again, I'll keep on updating,but very interesting point of
view about Apple's currentstatus from Apple to Google.
Google just announced a new coolfeature from Notebook lm, which
is public sharing.
You can now share your notebookswith anybody in the world, even

(43:18):
if they don't have a Googleaccount.
It means that one of the besttools in the world today to
summarize information and allowasking questions about
information, you can now make ita much more collaborative
environment.
Teachers can create interactivestudy guides for students.
Startups can build product hubsthat multiple people can
participate in and learn about.
Researchers can share findingsthrough that and business people

(43:40):
can share project data and so onwith other people.
The tool just became morecollaborative, which is great.
I use Notebook Lamb all thetime.
I think it's a fantastic way tosummarize information from
multiple sources, be able to askquestions about it, and so on.
And from Google to perplexityintroduced Perplexity Labs last
week, which we talked about,which is a really cool tool.
I still think it's behind themore agentic tools like Gens,

(44:03):
spark, and Manus, which have waymore advanced agentic
capabilities compared toperplexity Labs.
And if you wanna know more aboutGens, spark, and Manus and tools
like that and how to run themsafely, don't miss the next
episode that's coming out onTuesday.
This is exactly what we're gonnadive into, but staying on
perplexity, their CEO Arvin SWpredicts that AI agents will
redefine how we interact withthe web, moving beyond answering

(44:25):
questions to take actions likebooking rides, ordering foods,
and everything else that we doin the internet today.
Now Rivas aims to disruptcurrent search by building AI
agents that will integrateseamlessly with the apps.
And as we mentioned previously,they're coming up with a new
browser called Comet that willhave everything integrated into
it.
This browser is supposed to belaunched next month, so it's

(44:46):
just around the corner.
There are many other companiesright now that are building
quote unquote AI based browsers.
It'll be very interesting to seehow that works, but the
direction is clear.
We're gonna have a lot moreagent approach to interacting
with the web.
We're gonna see less and lesshuman traffic to websites as
time goes by, and companies whorely on human traffic going to

(45:06):
their website, whether fore-commerce or for information
needs to understand that this isgoing to change.
It's not gonna happen overnight,but over the next five years,
there's gonna be a very clearincrease in agent traffic to
data, not necessarily websites,and a very significant decrease
in human traffic to thesewebsites.
And that means that companieshave to start adjusting to that
and figuring out what that meansfrom data structure,

(45:28):
architecture, and other aspectsof the way they're engaging with
their customers.
Now just like the potential iOScollaboration that I mentioned
before, perplexity secured adeal to pre-install its AI on
Motorola's new razor phones andSVAs credits, Google antitrust
scrutiny for loosening grip onOEMs that are now allowing them
to install their search andother capabilities on the phones

(45:48):
instead of, or in parallel toGoogle's offerings.
And from perplexity to meta.
Meta just announced in theirshareholder meeting that they're
aiming to automate the entirecreation of ads by the end of
2026.
We share that with you a coupleof weeks ago, but now we got
more details on what the plan,but the plan is very extreme,

(46:09):
and I'm quoting Zuckerberg fromthe meeting in the not too
distant future.
We want to get to a world whereany business will be able to
just tell us what objectivethey're trying to achieve, like
selling something or getting anew customer, how much they're
willing to pay for each resultand connect their bank account.
And then we just do the rest forthem.
It means that you don't need anykind of agencies and that it

(46:32):
will optimize automatically downto the individual level.
It means you don't need agenciesbecause it will create the ads.
It will create the copy, it willcreate the text, it will create
the images, it will create thevideo.
It will create everything thatit needs.
It will optimize down to thepersonalization to a specific
person based on their interest,based on their hours, based on
how they consume, what theyclicked in the past and so on.

(46:52):
Stuff that no agency and nohuman can do right now, and it
will optimize for all thesethings.
You're obviously in a questionwhether you are allowing your
creative to be controlled by amachine, but if that machines
know how to achieve the resultsbetter than you can with your
creative, then there's noproblem.
Well, there is a few problems.
Problem number one.
That you lack control on themessaging that is gonna be

(47:12):
presented on behalf of yourcompany, which might be a
problem, especially if you havea brand name that you're trying
to preserve.
I'm sure they will have sometools to guardrail what it can
and cannot say, and what brandguidelines to follow and so on.
But that wasn't clear so far.
The other thing is that it'sgonna kill, I don't know how
many, but many, many, manydifferent agencies who
specialize today in doingexactly that, which is doing

(47:35):
social media advertising onbehalf of clients.
There's been similarannouncements already by Google
and TikTok saying that they'regonna follow similar concepts.
So the idea of an agency thatbuilds ad and distributes ads
free because they know how to doit better than you might be the
thing of the past.
Within a few years from now,staying on Meta, they just
announced that they're gonnaswitch most of their risk
assessment from Human Review toAI review on Facebook,

(47:58):
Instagram, and WhatsApp.
Whether that's good or bad, I'mnot a hundred percent sure
there's goods and bads in bothof it.
I'm sure AI can look at morestuff.
I'm sure AI can classify thingsbetter.
I'm sure AI is gonna miss thingsthat the humans would've caught,
and so I really hope that's notgonna come to bite us.
You know where, but for now,this is the direction that Meta
is taking.
By the way, similar to Microsoftthat we talked about before,

(48:20):
meta is also going through someserious restructuring in its
generative AI group splittinginto two units to address what
they call quote unquote internalchallenges.
So in the new leadershipstructure, Ahmed Al and Amir
Frankl will co-lead the new A GIFoundations team, focusing on
LAMA models and AI agents.
And Joel Pue, VP of AI researchwill oversee the team dedicated

(48:43):
to AI integration in META'Sconsumer apps like WhatsApp and
Instagram.
So one more the front end, onemore the back end if you want.
This actually makes sense to me.
If you go back to how we startedthis episode, we're talking
about the fact that theapplication layer is gonna be as
important and I think in thelong run, more important than
the underlying models, splittinginto two teams makes perfect
sense to me.

(49:04):
And now to some very interestingannouncements from smaller
startup and not just from theGiants.
Foley, which is a company thatgenerates phone agents, has just
achieved 99.2% conversationalaccuracy, surpassing the
previous king, which was OpenEyes Chachi PT four, all with
94.7%.
Now, the biggest difference wascutting response time and

(49:25):
latency by over 70%.
How did they do that?
Well, I'm sure they'reoptimizing the models like
everybody else, but they alsodid it through partnership with
Grok with the Q.
So grok that develops andprovides the most advanced
inference platform in the worldtoday.
So these are chips that are notoptimized for training AI like
the GPUs from Nvidia, butactually are specialized in

(49:46):
inference, meaning generatingtokens.
So they have what they call amulti LoRa hot swapping that
enables instant model switchingin real time without any
latency.
And they're also using MIT izeplatform to optimize the
performance of everything thatthey're doing.
And they were able to reduce theresponse time from 661
milliseconds, basically overhalf a second to 176

(50:08):
milliseconds, which is whathumans do.
It reduces dramatically theability of humans to actually
know they're talking to machinesbased on their own internal
research.
They're seeing that about 70% ofpeople cannot tell that they're
talking to AI when they'retalking to their platform.
That's obviously coming fromtheir own CEO.
So I don't know how crediblethis is, but I can tell you for

(50:30):
sure that these agents aregetting very, very good.
They're doing an incredible jobin providing customer service.
They're gonna get really good atdoing outbound sales and inbound
sales and a lot of other stuff,and that's another industry that
is at complete risk ofelimination, which is the
industry of the call center andcontact center.
I think five to 10 years fromnow, the concept of contact
centers were just seized toexist.

(50:50):
There's still gonna be peopledoing maybe higher level more
relationship kind of things, orbeing supervisor to assist in
stuff that the AI wasn't able tosolve.
But that's gonna be singledigits percentages of the amount
of people that are currentlyemployed in the call center
industry.
Another company that made a hugerelease this week is Hagen.
Hagen is a platform that enablesto create human-like avatars for

(51:11):
any need that you have, whetherit's training, customer service,
onboarding, et cetera.
I use Hagen all the time.
I teach Hagen in my courses.
It's a fantastic platform.
Well they just launched AIStudio, which allows to take the
creation of the videos and therealism of the avatars to a
completely different level.
It includes multiple components.
Most of them are prompt based,so you can prompt how the video

(51:32):
will look like.
But they have now newfunctionality like voice
director that allows a muchbetter fine tuned avatar speech,
including the ability to dovoice mirroring, meaning you can
record your own voice and thenit captures the nuances of how
you speak.
And then you can apply thosenuances to any voice that you
can pick from their platform.
They're also allowing you moregesture control of how the

(51:54):
actual avatar moves their handsand what kind of gestures
they're going to have.
And they're going to bereleasing additional
capabilities like prom, controlover camera motion elements, and
prompt based editing andstreamline, including B-roll
adding all of that with a lot ofAI capability.
As I mentioned before, I reallylike Hagen.
I really like their offering andI'm very excited to test this
out.
And I promised you in thebeginning that there's gonna be

(52:15):
some additional news when itcomes to the scientific benefits
of ai, well, the big one is thatthe FDA just approved a tool
called Clarity Breast, which isa breast cancer detecting
platform.
So the Breast Cancer ResearchFoundation just announced the
authorization of Clarity Breastto be used with actual patients.
That's a big milestone for AIbecause it is the first AI

(52:37):
platform that will be used forsimilar things in the public.
It can predict a five year riskfor breast cancer using only
standard mammograms.
So today, the way breast cancerrisk is analyzed is based on a
mammogram with the human lookingat it combined with other risks
such as family history.
Well, this tool actually looksat very small nuance changes in

(53:00):
mammograms.
And using that because it has somuch information in the past, it
can make a much betterprediction of future risk.
This is fantastic news for womenand it's fantastic news for the
research of cancer becausesimilar things might be applied
to other ways of predictingcancer.
And as we all know, catchingcancer when it's in the early

(53:21):
stages dramatically increasesthe chances of a successful
recovery.
And so I find this to be reallyamazing and great news.
And in another breakthrough in afield that is less relevant to
most of us, but it's still veryinteresting.
New AI techniques allowsresearchers now to have better
analysis of cosmo logicalparameters and allowing them to

(53:41):
better predict and understandhow the universe works.
These new tools were used withhuge amounts of data from Sloan
Digital Sky Survey Boss Dataset,it allows researchers to
identify subtle patterns thatwere invisible in traditional
methods and helping increase theaccuracy of predictions and
information by 30%.

(54:02):
This is a huge increase.
Again, it doesn't have any dailyapplication to us, but if you
take that to the world ofresearch and you combine these
two last pieces of news, itshows you that AI's ability to
look at huge amounts of data,whether visual, numerical or
other, and make sense in thatdata, is gonna allow us to drive
significant new innovations andresearch in many fields, which
is really exciting.

(54:23):
That's it for today.
We'll be back on Tuesday.
As I mentioned, talking aboutgeneral agents like Manus and
Spar and how to use them safely.
so I highly recommend you checkthat out.
I think it's gonna open youreyes to what's possible today,
which most people do not know.
Quick reminder for the coursethat is coming up in August.
So if you're looking to take ourAI Business Transformation
course, you should check out thelink in the show notes right

(54:43):
now.
And if you're enjoying thispodcast, please share it with
other people.
You can do this right now.
When you're done, open yourphone, click on the share button
and share it with four to fivepeople that can benefit from
listening to this podcast.
You are helping drive AIliteracy and I also be very
grateful if you do that.
And until next time, have anawesome weekend.

All Episodes

195 | The AI business application battle is intensifying, Self improving AI, AI for early cancer detection, and more AI news you should know for the week ending on June 6, 2024

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

All Episodes

Stuff You Should Know