# AI Shifts: Open Models, Solopreneur Boom, and Compute Risks

**Podcast:** The AI Daily Brief (Formerly The AI Breakdown): Artificial Intelligence News and Analysis
**Published:** 2026-07-06

## Transcript

Today on the AI Daily Brief, the data is in and AI seems to be changing the nature of entrepreneurship.
Before that on the headlines, the CEO of Palantir says the government is turning towards open-weight models.
The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.
All right, friends, quick announcements before we dive in.
First of all, thank you to today's sponsors, KPMG, Retool, Blitzy, and Airtable.
To get an ad-free version of the show, go to patreon.com slash ai daily brief, or you can subscribe to Apple Podcasts.
And if you want to learn more about sponsoring the show, head on over to ai daily brief dot AI slash sponsors or send us a note at sponsors at ai daily brief dot AI.
We are kicking off this week with a number of stories that continue and drive forward the major themes from the past several weeks.
Last week was punctuated by a fiery rant from Palantir CEO Alex Karp during an appearance on CNBC.
He said that some U.S.
government customers are migrating to open source after AI sovereignty concerns.
In the interview, he said, Taking aim at the consulting spinoffs from OpenAI and Anthropic, he continued, Customers are not interested in some fake deploy code that transfers the alpha to a third party.
Karp suggested it's time to ask some hard questions at the frontier model companies like, Who owns the data?
Where is it cached?
Are the prompts secure if this was being transferred to you?
If it was so valuable and I can make you a billion dollars, wouldn't I say I'll make you a billion dollars and I want 30%?
Why are they charging for tokens if it's so valuable?
It was a fairly full-throated attack on Anthropic and OpenAI, with Karp arguing that data security should now be front of mind for many AI users, and a claim which obviously underpinned the business model of the whole diatribe, that open-weight models are now at the point where they can replicate the performance of proprietary models while minimizing that risk.
He said, we can take an open model and get it to the point of a frontier model, but you control the weights.
Doubling down in a follow-up interview with the information, Karp claimed that some government departments had already made the switch, saying that they're now using NVIDIA's open-source model Nemotron instead of proprietary models trained by Anthropic or OpenAI.
Karp said, there's just very deep frustration around, are they going to optimize the models for me, or are they going to take the alpha of my business, transfer in their weights, and compete against me?
Karp said that Nemotron is already providing, quote, equal or in some cases superior performance on the battlefield use cases, which are mostly highly classified.
Karp expects every Palantir client to begin using open models, quote, as soon as they see it being at parity.
Now, of course, both Anthropic and OpenAI have been very clear about their policies of not trading on enterprise customer data.
There was enough chatter following this interview that Colin Jarvis, who leads FTE efforts at OpenAI, had to tweet, At no point do we train on customer data.
We push the limits of the models to get our customers to production success and oftentimes share insights from our FTE's experiences in the field back into the organization.
Former AI czar David Sachs pointed to Anthropics' launch of Claude Design shortly after partnering with Figma as evidence that frontier model companies are more than willing to compete with their customers if it seems like the right business move.
Now, Palantir's carp has a very particular style, and I'm not really interested in litigating that.
What's interesting here is the potential mainstreaming of the open-weight alternatives are viable kind of argument.
Now, one interview alone isn't going to get people to necessarily change their minds or think differently about that, but this is the type of discourse that is going to have everyone from enterprise buying leads to Wall Street investors sitting up and paying attention.
And it shows, if nothing else, just how much more open the playing field looks like in this new token scarcity, token efficiency sort of era.
Yet when it comes to Wall Street and NVIDIA, This was not really the thing that people were talking about.
Instead, the story that's been making headlines for the last couple of days is NVIDIA backstopping AI demand in a bid to push NeoCloud growth.
NVIDIA announced the move in terms of a new business model, where the chipmaker provides guaranteed demand in exchange for a cut of revenue.
In a blog post announcing the new strategy, NVIDIA wrote, Emerging AI companies historically have had limited access to capital-intensive infrastructure, with even long-term commitments insufficient to unlock financing for compute.
To address this, NVIDIA is introducing a new business model that opens up compute access to the fast-growing AI ecosystem of startups, model builders, enterprises, research organizations, and regional AI players.
Now, in concrete terms, the deal will see NVIDIA renting back unused GPUs at a guaranteed rate if the NeoClouds fail to find demand in the market.
In exchange, NVIDIA will take a cut of revenue for all GPU rentals.
Now, of course, this is not the first time NVIDIA has backstopped AI demand.
Last year, NVIDIA signed similar deals with CoreWeave and Lambda, guaranteeing to buy any unused capacity.
However, the new model seems to be aimed at supporting smaller and less established companies.
The first two neoclouds taking advantage of the program are Firmus and Sharon AI.
Firmus is deploying a cluster of 170,000 GPUs in Indonesia, which is one of the largest data center projects planned for the nation, while Sharon AI aims to deploy 40,000 leading-edge GB300 GPUs.
Now, the critique of this sort of deal in the past...
has been that it has echoes of the vendor financing boom that came undone during the dot-com bubble.
Rich Dupree of 24-7 Wall Street, however, tried to explain why in his and many others' estimation this time is different, writing, In other words, what NVIDIA is doing...
is trying to provide the anchor demand that allows these projects to access financing from other sources.
Yes, it's a directional bet on rising AI demand, but NVIDIA isn't directly financing their own hardware or taking on the default risk of the neoclouds.
Also, we're talking about a backstop of a few billion dollars in an era where NVIDIA is generating $80 billion in revenue each quarter.
Still, this is one of those stories where how Wall Street decides to digest it is going to tell us a lot about the mindset and the narrative moment when it comes to AI investors.
Speaking of NeoCloud, SoftBank is launching a NeoCloud business to service U.S.
demand.
Called SB Neo, the new company plans to begin renting out their AI compute in April using infrastructure currently under construction.
SoftBank says they plan to scale to 10 gigawatts of U.S.
capacity by mid-2028 and also have plans to build gigawatt-scale data centers in Japan.
Now, the big question is what this means for SoftBank's collaboration with OpenAI.
The 10 gigawatts of U.S.
capacity likely refers to a campus in Pike County, Ohio.
currently under development by SoftBank subsidiary SB Energy in collaboration with the U.S.
Department of Energy.
Initially, reports suggested that OpenAI would lease this capacity upon completion, which could still be the case, but certainly SoftBank seems to be signaling that they want the site to be viable as an independent neocloud business as well.
Moving now from neoclouds to geopolitics, Alibaba has banned employees from using Claude over potential security risks.
Sources said the ban was announced on Friday and expected to come into force by the end of this week.
Now, Anthropic, of course, has already taken measures to prevent their technology from being used in China, but nothing that can't really be circumvented using a VPN.
In a report published in February, Anthropic accused the major Chinese AI labs of distilling their models with a level of specificity that implied that Anthropic were tracking the interactions.
Then last month, Anthropic wrote to Congress and accused Alibaba of brazenly, their word, carrying out a large-scale distillation attack on their models.
They claim to have uncovered a network of 25,000 fraudulent accounts that generated almost 29 million interactions with Claude, all tied back to Alibaba.
Meanwhile, Anthropik's account surveillance techniques were uncovered in a Reddit post last week, where the poster accused Anthropik of embedding spyware in Claude code which could detect the use of a VPN, and examined metadata to determine whether a user is in China and has ties to a particular lab.
Following that post, Anthropik's Tariq explained that they've developed stronger mitigation since then and will be removing the spyware.
So then we may be witnessing the success of those mitigations as Alibaba is apparently no longer willing to risk using Anthropics models.
In their memo to staff, Alibaba wrote, Claude Code was recently discovered to carry backdoor risks.
After comprehensive evaluation, Claude Code has now been added to a list of high-risk software with security vulnerabilities.
With Reuters said, Anthropics restrictions are difficult to enforce on individual users, but companies are more aware of legal and compliance risks.
There is a lot of back and forth and posturing and things going on behind the scenes here.
It's worth noting that Alibaba is trying to get removed from the Pentagon's blacklist, so maybe this is a part of that.
But regardless, it's pretty interesting stuff.
Lastly today, one that I'm sure we're going to come back to later in the week.
Tesla is the latest company to impose AI limits as token budgets get reined in.
The information reports that Tesla will limit employees to $200 per week in token spending beginning this week, although the new policy was announced last month.
Sources said that some software engineers at Tesla were routinely racking thousands of dollars in token costs each week.
Now importantly, The nuance here is that the policy isn't set in stone, with workers able to request a higher budget if necessary.
In fact, at some point this week, I think this is going to be worth getting into in a much deeper way, because of the nuance here that I think makes Tesla a much more interesting case study than it seems at first.
But for our purposes today for these headlines, it is, as I said at the beginning, another example of just how things are changing in this new period that we are entering into.
For now, however, that is going to do it for the headlines.
Next up, the main episode.
One of the most important AI questions right now isn't who's using AI, it's who's using it well.
KPMG and the University of Texas at Austin just analyzed 1.4 million real workplace AI interactions and found something surprising.
The highest impact users aren't better prompt engineers, they treat AI like a reasoning partner.
They frame problems, guide thinking, iterate, and push for better answers.
And the good news?
These behaviors are teachable at scale.
If you're trying to move from AI access to real capability, KPMG's research on sophisticated AI collaboration is worth your time.
Learn more at kpmg.com slash US slash sophisticated.
That's kpmg.com slash US slash sophisticated.
This episode is brought to you by Retool.
Generating a working app now takes about five minutes thanks to AI.
Getting it safely into production with auth, permissions, audit logs, security reviews, that part still takes time.
That's the gap Retool closes.
Build apps however you want.
natively in Retool with Claude Code, Codex, or any coding agent, or by importing React you've already built.
It all deploys into Retool and picks up governance automatically.
Security lives in the platform, not in whatever the AI wrote.
So your team ships at AI speed without the shadow IT and the endless reviews that stall out most vibe-coded projects.
It's why teams at Amazon, Stripe, and Brex build on Retool.
And new enterprise customers who sign by September 30th get up to $10,000 in AI credits per year.
Start building at retool.com slash AI daily.
With the emergence of AI code generation in 2022, NVIDIA master inventor and Harvard engineer Sid Paresci took a contrarian stance.
Inference time compute and agent orchestration, not pre-training, would be the key to unlocking high-quality AI-driven software development in the enterprise.
He believed the real breakthrough wasn't in how fast AI could generate code, but in how deeply it could reason to build enterprise-grade applications.
While the rest of the world focused on co-pilots, he architected something fundamentally different.
Blitzy, the first autonomous software development platform leveraging thousands of agents that is purpose-built for enterprise-scale codebases.
Fortune 500 leaders are unlocking 5x engineering velocity and delivering months of engineering work in a matter of days with Blitzy.
Transform the way you develop software.
Discover how at Blitzy.com.
That's B-L-I-T-Z-Y dot com.
This episode of the AI Daily Brief is brought to you by HyperAgent, where you run fleets of agents your team can manage together.
New users get $1,000 in inference.
Forget local agents and chat workflows waiting on your laptop to be prompted.
Hyperagent deploys always-on agents in the cloud, doing real work across the tools your team already uses.
Marketing's agent turns competitor moves into landing pages.
Sales's agent enriches leads, drafts emails, and updates the CRM.
Ops agent chases the paperwork and tracks the budget.
Every agent has access to shared context and follows your rules about scope and approvals.
It's time you add agents that feel like teammates.
Hire yours at Hyperagent, built by the team at Airtable.
Claim your $1,000 in inference at hyperagent.com slash AI Daily Brief.
Welcome back to the AI Daily Brief.
We are now past the 4th of July, which means by any and all accounting, we are fully in summer mode.
Now for some folks, especially in the corporate sector, that means naturally a little bit of a turning down of the intensity, a time to pause and catch one's breath, maybe take a vacation, or at least shift into a slightly different gear.
And yet for some, this is a moment to turn up.
And this weekend, the Wall Street Journal did a long-form piece about the changes in how some elite college students are thinking about their own careers, shifting away from traditional internship-type opportunities, and towards startups.
Now, we're going to expand that out today.
And so we're going to look at not only the anecdotes from the Wall Street Journal in that piece, but some broader data in how AI is shifting startups in general, but smaller, more nimble, solopreneur-type businesses in specific.
Now, when it comes to this piece, the Wall Street Journal wrote, For decades, the path of many elite students was clear.
Secure internships in tech, finance, or consulting, graduate with a cushy job, and climb the corporate ladder.
And yet, as they point out, given just how uncertain the job market seems to be, taking a chance as a founder or early startup employee is starting to seem a lot less risky than it might once have been.
The journal profiled several young students who have made their way to California to participate in the AI boom.
One subject of the piece was Princeton student Charles Muehlberger, who has taken a gap year to build a company that deploys local AI models.
Four weeks after touching down in San Francisco, Muleberger is headed to Barcelona to pitch his first customers.
And while the uncertainty of the traditional white-collar career is a big part of the story, there's also an inverse to that fear side, which is about the idea that AI is giving new opportunities to the most high-agency students.
Muleberger said, those who are building now get a voice in what the future looks like.
The journal also profiled a host of different programs that are offering free housing, mentorship, and networking opportunities specifically for students participating in the startup ecosystem.
One such program is the Yale Hacker House, backed by several alumni and associated venture firms.
Fifteen students are crammed into an apartment in Knob Hill looking to build the next big thing.
Another incubator covered is TechTrek, which says they aim to be a bridge between the startup and academic world, having recruited folks from MIT, Harvard, and Princeton, and saying that they're giving students an opportunity to take in summer in San Francisco before bringing that energy back to campus.
Now, one of the big debates surfaced in the piece is whether students should be dropping out of the Ivy League to build an AI startup.
Leah Ryan, one of the creators of the Yale Hacker House, is fully committed, walking away from a PhD in genetics and a job offer at a leading biotech firm.
Her startup called Cortex was founded in March and recently raised at a $10 million valuation after signing their first commercial client.
Ryan said, When you raise money, I think it's actually quite irresponsible to be in school.
Any serious founder will drop out.
Others see a degree as a necessary part of business success as well as an important safety net.
Gari Shetri, a Princeton sophomore and the founder of AI reporting startup Strata said, You kind of always want to have a degree at the end of the day.
Now, to some extent, There's nothing particularly novel about this conversation.
If you went back and looked over the course of the 20 years since I graduated college, you could probably find, with pretty frequent repetition, some version of this story of startups versus more traditional career paths.
The difference now, of course, is the extent to which AI is changing the calculus.
And that's on two different levels.
AI is impacting the activation cost of building a startup, basically lowering the barrier to entry, lowering the cost to actually build something meaningful and try to get it into the world.
But I think what's really important, and why I think you're probably going to see more articles like this coming up, is that AI is also upending people's assumptions about what safe actually looks like on the other side as well.
If the historic calculus is startups in solo endeavors as one side of the risk spectrum, i.e.
the very risky side of the spectrum, and traditional corporate paths as the less risky side of the spectrum, well, that starts to look a lot different when people aren't sure what the roles in the corporate world will even look like once AI is fully integrated.
And I think that rather than viewing startups as some unique and differentiated category, it's worth actually thinking about, in general, how they fit with how working is changing more broadly.
Economist Leah Palagashvili also had a piece in the Wall Street Journal this weekend called Me, Myself, and AI.
Tweeting the piece, she wrote, What if AI's first labor market effect isn't replacing workers, but making traditional firms less necessary?
And what's interesting about Leah's piece, Leah writes, Most debates over AI begin with the same question, which jobs will AI destroy?
But the first labor market shock may not be mass job loss, it may be worker migration from traditional firms.
AI is making it easier for one worker to do tasks that once required a small team.
In industries where AI can handle research, drafting, coding, editing, and analysis, the result isn't necessarily unemployment, it may be independence.
Now, Leah points out that this is actually starting to show up in the data as well.
She points out that since early 2024, Solo business applications, as tracked by the Census Bureau, have risen nearly 27% in professional services, information, education, finance, and insurance.
All sectors, she points out, have the highest AI adoption rates.
Compare that to areas like construction and wholesale trade, where solo business applications have been essentially flat.
She points out that if you go back to 2022 and 2023, the solo applications across all sectors were pretty similar, and that the divergence only emerged after 2024.
Leah argues that federal labor market data is painting pretty much the same picture.
Between 2022 and 2025, in occupations that were highly exposed to AI, solo self-employment rose about 20%.
In the least AI exposed applications, it didn't really change.
Taking one example, management analysts, which she argues is a useful proxy for consulting work, overall employment grew 12% from early 2022 to early 2026, while solo self-employment in that same occupation grew more than twice as fast.
A recent report from Stripe's economics team also highlighted this boom in solopreneurship, i.e.
solo operators generating significant revenue without hiring employees.
The report cited similar data from the Census Bureau, which classifies new business applications as either high-propensity employers or likely non-employers.
There has been a huge uptick in likely non-employers over the past 18 months, beginning in late 2024.
Now, Stripe notes that this signal has been misleading in the past.
The IRS has periodically pushed gig workers to complete business registrations over recent years, driving a fairly steady increase in non-employer business applications.
There was also the Paycheck Protection Program, or PPP, loans in 2020, which required a business registration and little else to qualify.
That policy was the clear driver for a doubling in non-employer registrations that year, which have remained elevated ever since.
This time around, however, we have a large surge in non-employer registrations while high-propensity registrations remain flat.
And while Stripe said that they can't distinguish between solopreneurs and new businesses with employees, they are seeing much faster revenue ramps for businesses that signed up in recent years.
They wrote, Businesses that signed up on Stripe after 2023 reached material transaction volumes earlier than the sign-up cohorts that preceded them.
The share of businesses, not just solopreneurs, reaching a million dollars in cumulative revenue within a year after going live on Stripe was roughly 30% higher for the 2025 cohort as it was for the 2023 cohort, and it was roughly 3x higher for the 2025 cohort than the 2019 cohort.
Stripe also noted a series of coincident indicators that suggest that this is a genuine boom in entrepreneurship rather than a quirk of the data.
New business registrations are up across a range of different jurisdictions, rising by 40% in Australia, 70% in Finland, and 80% in France since 2017.
Acceleration meaningfully increased in 2025 across each of these three countries.
Stripe also noted that we're seeing a boom in Delaware LLC and corporations up 40% year-over-year from early 2025.
Stripe also dug into revenue metrics and found that solopreneurs are doing better than ever.
They analyzed the data by creating a proxy index by screening for subscription to solopreneur-focused platforms.
And while the index isn't perfect, Stripe believes it's directionally correct.
And that direction is up and to the right.
The number of solopreneurs earning a million dollars more than doubled between 23 and 25.
Right, Stripe?
These trends are quite striking and suggest a true shift to the scale solopreneurs can reach on their own and the frequency with which they do so.
Now, Stripe's thesis, perhaps unsurprisingly, is that AI is enabling this boom in solopreneurship.
Now, this is not, they argue, a wave of vibe-coded apps hitting a million dollars in ARR.
but instead AI helping to fill the gaps that once made hiring mandatory.
For small businesses, they believe that AI services are now stepping in as the technical co-founder or the sales and marketing first hire.
Right, Stripe?
Part of the reason that businesses historically tended to be built by groups was that a single individual rarely possesses all the skills needed in the entrepreneurial journey.
Whether it's how to evaluate or size a market, code an app, price a product, write and execute a marketing campaign, or close a deal, AI can fill many of the gaps that founders previously turned to another human for.
The other driving force as Stripe presents it is the acceleration of sales activity from AI funnels.
Stripe wrote that AI-influenced user journeys now represent four times the share of new signups for their service, and they imagine the same is true across the board.
Essentially, if you're a solopreneur offering a great product, ChatGPT itself is recommending you and driving a healthy portion of your sales.
Stripe concludes, We think this phenomenon is the true engine of the AI surge in business formation we're seeing today.
The availability of this breadth of on-tap assistance allows anyone with sufficient motivation to go it alone.
Separately but relatedly, Stripe Atlas, which is the company's startup enablement suite, also recently discussed why, as they put it, solo founding is at an all-time high.
Now, this isn't just solopreneurs.
This is all sorts of companies, including those with big ambition and venture-backed goals, who are starting with a single founder instead of a set of co-founders.
Back in May, Atlas wrote, Solo startup founders account for 63% of C-Corps formed so far in the second quarter of 2026, an all-time high.
As more founders start companies on their own, the gap between typical companies and top performers is widening.
Now from there, Atlas dug into what makes these companies different, pointing out that they are often building AI-native products, selling globally from launch, focusing on B2B use cases, and winning higher customer retention from the start.
Former Atlantic journalist and author Derek Thompson Recently wrote a blog post about all of this called There's Never Been a Better Time to Get Rich Working Alone.
In a tweet he summed up, the debate about AI and jobs often breaks down into two extreme groups, both of which have an evidence problem.
On the one hand are the doomers who say AI will take everyone's job, even though unemployment remains low and the employment rate for prime age Americans is still very high.
On the other are the deniers whose insistence that AI is a worthless scam prevents them from seeing the many ways it's changing work in the economy already.
Now, if you want a strong and evidence-based take about AI and jobs, I have one for you.
there's never been a better time for workers to get rich by going independent.
This is a golden age for tiny startups with big revenue.
Now Derek points to all of the same data that we've just talked about and starts to explore what some of the bigger implications might be.
He wonders if more solopreneurs will create more aloneness and wonders what impact it might have on tax revenue for the government.
But the point is that increasingly, in a sea of speculation about potential future impacts of AI on work, The one area where there is starting to be very clear data right from the beginning is around solopreneurship and startup and small business formation.
And one area that I think is worth keeping an eye on is not just solopreneurship specifically, but also how the smallness and streamlining finds its way into other types of organizations as well.
A recent study out of Harvard Business and Inseed did find that in fact, AI-powered startups are running leaner.
SSRN summed up the study's findings.
AI-native startups are 25% smaller, flatter, and more engineer-heavy, yet equally valued.
Embedding AI into the product lets them scale knowledge work without large teams.
And this is, I think, why it's worth paying attention to these trends, even if you are not a solopreneur or startup person yourself.
What startups, and especially solopreneurs, represent is the extreme tail of the efficiency gains possible from AI.
They are going to speedrun a lot of the experiments that will eventually find their ways into other types of organizations.
And so to the extent that the data is showing that these companies are getting more successful faster with fewer resource inputs, it's highly likely that some parts of that find their way to other types of business environments as well.
Anyway, I think it's encouraging that we're starting to actually be able to debate AI's impact on the basis of data, not just on speculation.
And honestly, for solopreneurs, the numbers are looking good.
For now, that's going to do it for today's AI Daily Brief.
Appreciate you listening or watching as always.
And until next time, peace.