Free with a sign-in.
The best 16 min of 59
Factory’s Matan Grinberg on Recursive Self-Improvement, Chinese Models & Lessons From String Theory
Matan Grinberg says AI is making software engineering autonomous—and shifting the scarce skill from writing code to deciding what to build.
Sabrina Halper Show · Sabrina Halper
- Matan Grinberg is cofounder of Factory.
The best minutes
- 23:56Enterprise AI is shifting from usage targets to ROIPro pick 5 min
- 28:05Open models can be worth the risk of a closed monopoly6 min
- 40:53Recursive self-improvement does not guarantee runaway growth6 min
Every moment plays free. PRO marks a write-up that's in Pro.
Also worth watching
- 18:14A two-year COBOL migration took four days with a Droid5 min
- 49:03Choosing what not to build is the new software moat3 min
- 35:19The proposed policy exempts open models from federal review4 min
- 38:37Agents could cause a billion-dollar security incident this decade2 min
- 8:54String theory taught Grinberg to decide with incomplete information3 min
Skip the opening angel-investor story, childhood anecdotes, most of the string-theory primer, and the closing investor and advisor roll call.
Full analysis · 5 min read
Real Alpha · Nov 8
Matan Grinberg says Droid can migrate COBOL in four days
- Grinberg said Factory is seeing enterprises do roughly 30-year-old COBOL migrations in four days, versus two years before.
- The example lacks scope and measures, and Factory sells engineering agents.
- Leaders should measure accepted work, cost, and risk, not token use alone.
Free with an account
Read the rest of the analysis
7 sections, the 7 other voices it tests Grinberg against, and what it means for you. Every episode we cut gets one.
Check your inbox
We sent a link to your address. Click it to finish signing up. Not there in a minute? Look in spam.
or
Signed up before? The same buttons sign you in. The episode, its moments and the transcript stay free.
A four-day migration needs scope before it sets a benchmark
Grinberg said Factory is seeing enterprises do what he described as roughly 30-year-old COBOL migrations. He said work that used to take two years can now take four days with a Droid.
That is striking, but the episode offers no disclosed benchmark. Grinberg did not name customers or give codebase size, acceptance criteria, human review hours, or total cost.
The commercial context belongs alongside the result. Grinberg co-founded Factory, whose Droid agent handles engineering tasks, and Halper disclosed she is a small Factory angel investor.
“they can do that in four days with a droid just working on its own”
Our read
Treat this as evidence that agents may make legacy work viable, not as a planning estimate. Ask for scope, defects, review hours, and cost before putting four days in a forecast.
Choosing work is becoming the engineering bottleneck
Grinberg says the bottleneck is increasingly deciding what work to delegate, not getting an agent to execute. More output can still mean low-value output.
Developers still need to provide context and priorities. He says Droid can now figure out more about past work and likely future tasks.
For example, he points to a person deciding which tasks a Droid should prioritize. The migration story does not disclose setup or review time.
Our read
Track time from request to accepted result, including setup, corrections, and oversight. Faster code generation is not the same as faster delivery.
Factory now values agency over coding speed
Factory's hiring signal follows that bottleneck. Grinberg says coding contests and language fluency matter less, while high agency matters more.
He names former founders as the strongest predictor of success at Factory. That is an internal observation, not validated hiring research.
The episode gives no sample size, comparison group, or role breakdown. Eric Schmidt said agents could flatten business on In Good Company with Nicolai Tangen in October. He also said most large companies will struggle to adapt.
Our read
Test whether candidates can own an outcome, surface missing context, and verify results. Track job performance before treating Factory's preference as a general hiring rule.
Token budgets need returns, not usage quotas
Grinberg's enterprise prescription is to replace broad usage targets with return per token. He describes board pressure, then token-maxing, then allocation by incremental value.
He allows temporary overspending to break old habits, but not as a permanent measure. The framework asks whether one more dollar of compute beats a hire or a cheaper model.
But he does not define ROI, attribution windows, quality floors, or error costs. For example, Grinberg says enterprise sales may need cheap summaries, while engineering and data science can benefit from more tokens.
Aghi Marietti said on MTS in October that they were running about 500 trillion tokens, with volume growing 140x year over year. That describes his company's volume, not every enterprise.
Arvind Krishna asked on All-In in October which enterprise would reject a 10-times cheaper answer. In his specific case, a 40-times cheaper answer could do effectively 95% of the work. Guillermo Rauch said on Sourcery in October that open models handled 80% of Vercel's token traffic.
These accounts show adoption and possible price differences, not a way to calculate Factory's proposed ROI.
- AMAghi Marietti · MTS · Oct 23Enterprise AI is spreading across models as token use surges
Arvind Krishna · All-In · Oct 21Open-weight models could handle most enterprise AI at lower cost
Guillermo Rauch · Sourcery · Oct 16Open models now handle most token traffic on Vercel
Our read
Set budgets by task and compare incremental cost with measurable outcomes. Usage growth and cheaper answers are inputs, not substitutes for a company's own quality and return measures.
Open models hedge monopoly risk but add exposure
Grinberg's tradeoff is explicit. He would accept risks from Chinese open models rather than leave capable systems only with closed providers.
That is a competition argument, not a security finding. He describes trigger-based data exfiltration and subtle vulnerabilities in generated code as possible risks, not verified incidents.
He says US data-center hosting would not eliminate every risk. Factory routes among closed and open models, and Droid is its engineering agent.
Factory sells an engineering agent and routes among providers. Grinberg has a business interest in keeping model options available.
Our read
Treat openness as a competition choice, not a safety certification. Before expanding use, compare code defects, data handling, and auditability. Hosting locally does not by itself test hidden triggers.
Grinberg separates self-improvement from runaway growth
Grinberg distinguishes a model helping improve itself from a system accelerating without human help. Current contributions to model improvement do not establish hyperbolic growth, he said.
He has not seen signs that people can simply let go. His post-AGI label is his own definition, meaning AI already lets people tackle problems they could not solve before.
He says models have advanced math by finding counterexamples and extending thresholds, but remain weaker at defining wholly new theories.
Richard Socher described agents beating community benchmark results after days on the TWIML AI Podcast in November. That shows rapid search on defined tasks, not an autonomous research lab.
Qasim Wani said on MTS in November that RSI requires agents to manage data, experiments, evaluations, and infrastructure. Herbie Bradley said on MTS in November that fast takeoff remained a low-probability scenario.
Richard Socher · The TWIML AI Podcast · Nov 3Recursive reports beating community results on AI benchmarks- QWQasim Wani · MTS · Nov 6Recursive self-improvement means inventing a new lab
- HBHerbie Bradley · MTS · Nov 7Fast takeoff remains a low-probability scenario
Our read
Budget for faster iteration, but set a separate evidence bar for self-directed research. Benchmark gains alone do not establish an autonomous improvement loop.
Cheap software raises the value of saying no
Grinberg says cheaper software shifts advantage toward product judgment and choosing what to exclude. He says technical exclusivity no longer reliably separates software winners.
That is a strategic view, not proof that every product is interchangeable. Grinberg's point is about choosing what to build, not a claim that technical judgment disappears.
For instance, Grinberg says a consumer packaged goods company probably does not need its own coding agent. He argues teams should focus on their core strengths and use tools that do not distinguish them.
Factory applies that discipline internally, he said, by rejecting work that might be a passing fad.
Our read
Before an internal build, name the distinct user problem, the team's right to win, and the owner of ongoing support. Cheaper prototypes may still carry maintenance and distraction costs, so include those in the decision.
What it means for you
- Log accepted work. For each agent task, record completion time, review hours, correction rate, and whether the output reaches production. Compare these with the manual baseline before using the four-day example in plans.
- Set task-level budgets. Compare the next dollar of compute with hiring, a cheaper model, or leaving the task manual. Separate relationship-led work from engineering and data tasks, as Grinberg suggested.
- Test model exposure. Run code review, data-egress, and trigger-focused tests before expanding access to open models. Track defects and data flows, since local hosting alone does not address every risk.
- Rework hiring screens. Give candidates an ambiguous workflow and ask them to set priorities, choose tools, and define acceptance checks. Compare outcome ownership with speed on language-specific exercises.
What to watch
- Whether migration reports include scope, accepted quality, review time, and full cost, not only elapsed days.
- Whether agent-heavy teams report security testing and incidents alongside capability gains. Grinberg relayed Eno's prediction of a billion-dollar agent incident this decade.
- Whether self-improvement evidence covers autonomous research across data, experiments, and evaluation, rather than benchmark gains alone.
Send this to engineering leaders deciding whether to automate legacy work or add AI budgets.
Watch the minute where Grinberg gives the migration comparison
Written by Real Alpha from the episode's transcript and the other moments it links, then checked against the transcript. Quotes are the speaker's exact words.
Minute by minute
A Pro write-up, open to you here. Try Pro free for 7 days
A Pro write-up. The minute by minute is in Real Alpha Pro; the episode, its moments and the transcript stay free.
Read it with Pro, 7 days free- 3:09Math set Grinberg on a physics path
Grinberg says he moved around the South Bay 10 or 11 times as a child, got in trouble at school, and excelled at math. After his geometry teacher told him to retake the subject, he studied ahead and chose string theory after asking his father what the hardest math was.
- 10:57Physics taught him to judge with limited information
At Berkeley, Grinberg decided he did not want an academic career, which he saw as centered on teaching. After considering quantitative careers and taking computer science courses, a program-synthesis seminar drew him to code generation. He saw machine learning for writing code as a fundamental problem.
- 17:06Factory aims to automate engineering work
Grinberg says he cold-emailed former physicist Shaun Maguire after deciding a company was the best way to pursue code generation. Factory’s goal from the start was autonomous, not just assisted, software engineering. He says better models and developers’ willingness to rethink workflows have made Droid more useful; one enterprise used it to complete a 30-year-old COBOL migration in four days instead of two years. He sees prioritizing work and high agency, rather than coding speed, as key constraints.
- 23:15Enterprise AI needs ROI measurement
Grinberg describes an adoption path from board pressure to broad usage and then to measuring return on investment. He says temporary overspending can help break old habits, but companies should assess the return from each additional token and allocate accordingly. Relationship-driven sales may need cheaper models, while engineering, product marketing and data science can benefit more from additional tokens.
- 28:33Open models limit closed-provider dependence
Grinberg says open models are needed to prevent closed providers from controlling token supply, and he would choose Chinese open models over a world with no capable open models. He prefers frontier open models from the United States, arguing that model use could subtly shape users’ values. He says Chinese models could pose hard-to-detect risks, including triggered data exfiltration or subtly vulnerable code, but considers those risks preferable to a closed-provider monopoly.
- 33:30Security needs clearer accountability
Grinberg criticizes labs for presenting sandbox failures and unauthorized internet access as model capabilities, arguing those are deployment failures. He says the White House policy described in the episode would regulate US closed-model releases while allowing US open models to operate without that review. He says distillation gives non-US providers an advantage, and cites cofounder Eno’s prediction that agents could cause a billion-dollar incident this decade.
- 40:53Self-improvement may not mean takeoff
Grinberg says recursive self-improvement means models can improve themselves, but does not guarantee faster-than-human progress or runaway growth. He says models already help improve models, but he has not seen evidence that people can hand off the process and trigger hyperbolic growth. He calls the present period post-AGI because AI lets people tackle problems they could not solve before, and cites recent progress on mathematical problems as an example.
- 46:38Software abundance makes focus more valuable
Grinberg expects cheaper engineering to expand both the software ordinary people can build and the quality users expect, reducing friction in services such as DMV interactions. But he is not convinced every person or company should build custom tools; he says they should focus on their core competencies. With no software capability remaining exclusive, he sees choosing what not to build as a key differentiator.
- 52:15Loyal supporters matter more than logos
Grinberg says loyal backers matter more than prestigious names because investors cannot build the company for founders. He credits Shaun Maguire and Sequoia, and names Snowflake CRO Chris Degnan as an advisor who taught him enterprise sales, which he says is essential to building a strong business. He also praises Nikesh Arora and Chris Fredrickson for their advice and support.
Published Sep 2, 2026.
Transcript · 53 min read
Sabrina Halper Show, hosted by Sabrina Halper, with Matan Grinberg.
Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.
We are post-AGI right now. Like, we have AGI. The only world in which there's, like, there's nothing left for humans to do is if we solve all problems. But my sense is that the number of problems that exist is, like, here, and the number of problems that we're solving is, like, this. There used to be a moat in software of, we know how to build this and no one else does, so that's why we're going to win. That doesn't exist. There is nothing that nobody else can build, which is crazy. People talk a lot about recursive self-improvement. The idea that the models get so good that they start training themselves, and there's this loop where they just get, you know, insanely good but improve themselves is not synonymous with like hyperbolic growth it could be the case that we get to rsi and it's just never as good as human assisted improvement on these things what happened to a world where if there is a security incident you are like deeply ashamed and responsible for it not like bragging about it in a research blog post no open models or open models that are from china i will take open models from china for sure because the threat of the only good models being from closed providers is ultimate monopoly, like the monopoly to end
all monopolies. It's the interest of the entire economy for that not to be the case. Mouton Grinberg, welcome to my show. Thank you for having me. I am so excited you're here. And a disclosure is that you are the first person I've had on the podcast that I also have the honor of being a very small angel investor in. Too small. We can change that. and I actually when I was preparing for this episode I went back um to see like what was the first message I ever sent you and I like remember the whole story which is that I was like a year and a half out of college I was working in a venture fund in New York I was using LinkedIn sales navigator to like find people that I thought had like amazing backgrounds who seemed like they were doing nothing on LinkedIn which means like probably starting a company and I came across your profile. And I was like, who is this guy that working on a PhD in string theory has done all this like crazy physics work. And it's just like for the moment doing nothing. Went to your Twitter, it literally just said factory. And I was like, what is he doing? So I messaged you, I think this was like, May of 2023. And I was like, hey, Matan, you have an amazing background. I would be really
interested to learn more about factory. And that was the beginning of how we met. And the rest is history i think that was like a month after you had officially started that that is by far by the way like the first like investor inbound that we got we i think we were two it was me and my co-founder you know maybe our first employee had started or was like about to start but that was crazy i remember thinking how on earth does she like it was like i changed my twitter bio and within you know an hour or so it was that was crazy so props i was like fbi but i honestly like the first time we got on a zoom, I was so impressed with like your energy and your vision for factory. And I, I actually looked at all of these things and I was texting one of my best friends who was also an investor. And I was like, I just think Matana is going to be so successful.
Read the whole transcript
Like he has got the sauce. So like sitting here three years later, it's really so exciting to see how far you've come and like everything has kind of come to fruition and we're going to get into all of it, but I would really like to start from the very beginning of your story. Matan, who are you as a kid? Who is young Matan? Young Matan, that's funny. Young Matan grew up in the Bay Area. So like in the South Bay, my family moved around a lot. Like I probably moved, I think like 10 or 11 times before I went to college. And why did you move so much? It's so funny. So my parents were immigrants from the Soviet Union. They came from Moscow and Moldova um and they could never decide if they wanted to be closer to my mom's work or my dad's work and so my dad was in like South South Bay my mom was uh she worked at Stanford Children's Hospital so she was around Palo Alto um so we kept like alternating between like Saratoga and like Palo Alto and then supertino and then los altos and so moved around a bunch between their power struggle yeah yeah um and uh and it was also you know their like immigrant mentality of like they were trying to you know see if along the way they could like you know uh if they say you know sometimes they
would buy a house and try to you know upsell it after a year or something uh sometimes they would just rent but um so that was kind of the background there um as a kid i was very much a troublemaker Like, I would get into so much trouble in elementary school and middle school. I got suspended so many times. I was also really good at math. But I think, you know, retrospectively, I think I was, like, bored by things. And I thought I was, like, too smart for stuff. So I would, like, skip class and, like, go and just, you know, wreak havoc. Were your, like, immigrant parents from, like, the Soviet Union, I imagine that they really valued education.
Like, how did they feel about that phase of your life? I mean, it was like, you know, it was a huge disappointment. Like every summer was like math camp or like physics camp or, you know, some some combination of those. So they were very much doing the right thing there. And then it was just like they would get a call from the principal being like, you know, here's what Matan did again. So they were very, you know, displeased, I would say, or they were concerned. They were worried if they were doing something wrong. uh my part it makes me feel bad thinking back because like they i just put them through so much it all worked out i'm sure they're they're very proud now um what set you on the physics track and and and what about string theory hooked you yeah yeah so um so funny enough um so in eighth grade so like was about to transition into high school um i still like was not a particularly good student but i was really good at math and i like i knew i was good at math and I you know probably had some ego about it and then my eighth grade geometry teacher at the end of the year like while everyone was preparing to go to high school she was like oh yeah Matan you need to retake geometry in high school and I like I was so offended and pissed off by that and it was
like she thinks I need to retake geometry like I know all this stuff and so out of spite my first order on Amazon even like today like if I open up my Amazon app and just scroll all the way down my first order on amazon was textbooks for algebra two trigonometry pre-calculus like calc one two three you know differential equations and you know i was basically like okay over the summer like i'm gonna show her like i'm gonna study all of these textbooks i'm gonna place out of you know all these classes and then i asked my dad what the hardest math was and he said string theory and i was like okay i'm gonna be a string theorist and it turns out i mean string theory is theoretical physics not exactly math but you know close enough um and so you know i ended up just spending the whole summer doing that um i went to a very small uh private school in uh south bay that no one has ever heard of like my graduating class with like 10 people um but the one thing that was great there is that they allowed me to take online classes for credit and so i'd like placed out of all those math classes and it was like okay physics is like the only thing i care about like i want to become the best physicist in the world wow um so it's kind of out of spite uh was the
was the origin but it actually turns out that the mathematics of string theory is probably like the most one of the most beautiful artifacts that humanity has created in what way like i guess taking a step back like what even is string theory so yeah what is strength yes so okay so yeah so okay so you're like talking about the summer of math textbooks it genuinely sounds like my worst nightmare no it's okay so so the two most predictive theories humanity has ever come up with so predictive theory right like if we want to you know roll a ball down a hill and predict exactly the dynamics of what's going to happen um but you know for broader much more broad physical phenomena the two most predictive theories humanity has ever come up with um are quantum field theory which is like the physics of the very small so particles electrons photons you know all that stuff the physics of the very small quantum Thank you.
theory and then the physics of the very large things general relativity so that's like galaxies you know like all the all the big stuff um both of these theories we can test their predictions and it's up to like 13 to 15 decimal places of accuracy is how strongly predictive those theories are which is like insane like it is so so precise um and so that's fantastic it's like okay these must be universal truths because yeah you can predict all these little dynamics so well and so So, you know, physicists naively thought, okay, well, if these are two of the best predictive theories we've ever come up with, surely they're both true and they work together and, you know, physics is done. But it turns out that if you try to bring these two theories together, everything breaks. Now, you might ask, where do these theories come together? Because one is the very small, one is the, you know, very large. Why is that even relevant? Because, you know, they don't come together. Well, it turns out that black holes have singularities that are infinitesimally small but extremely heavy and so there you need both general relativity and quantum field theory and it just so happens that when you try to merge them like everything breaks all the
“good judgment with minimal information because at the end of the day like to write that paper i didn't become”
where you're like fuck this yeah well so okay so the paper so the paper with juan was uh at my undergrad um and then i came to berkeley just over there uh to do uh to do my phd and at berkeley is when i realized that i didn't want to do this for the rest of my life in particular because i realize and I knew this the whole time but it became viscerally true that wait like being to do theoretical physics you have to be in academia and being in academia means you're a glorified teacher yeah and I'm not a particularly good teacher um and I didn't get a lot of excitement
out of it and I was like oh my god the rest of my life is like just going to be surrounded by 18 year olds yeah because like I already know everyone the strength of your community is very Christmas. Also, I just to ask the elephant in the room is like, is there anything going on with string theory anymore? People are saying it's dead. Like there's, there's definitely a lot of exciting stuff. Now, is it the most, you know, fast moving? Yeah, probably not. I think. And I think, you know, theoretical physicists are generally really smart. And a lot of them over the last decade have been leaving theoretical physics. I mean, there's a reason I think like half the co founders of Anthropic were theoretical physicists.
Jared Kaplan was, like, exactly in the same, you know, area of physics as I was. Dario was a physics back in the day, also at Princeton, but he was more on the bio side. But, like, there's been an exodus of physicists to, like, more interesting problems, and I think that's for a good reason. Okay, so once you decide that you don't want to spend your life in academia studying theoretical physics, what do you do? so it was honestly like it was a big existential crisis because like my whole identity was like physicists have with them kind of an arrogance sometimes which you know i'm probably i fell victim to that a little bit where you're a confident guy but there was like there's a quote from um uh from uh what's his name i think it was i think it was earnest rutherford okay i believe it was yeah we can yeah we can like talk track over if it's not rutherford um but it's basically he says all of science is either physics or stamp collecting which is basically like everything at the end of the day is physics or it's just like labeling stuff and you know it's kind of irrelevant um and so there is that kind of arrogance associated with physics and i kind of built my whole identity was i'm a physicist and then kind of all came crashing down because i was like
I don't want to do this and I'd spent the last decade only caring about math and physics all day like that was all I did and you never held a job right never like you weren't doing some I've literally I'd literally never been paid for work except for like physics research when they would give grants um and so it was big existential crisis and it was like you know what okay like before I figure out what to do like I should probably become like a normal human and like read literature and so I went and like looked up like the top 100 pieces of literature and just like read all of those or went to the imdb top 250 movies and just like watched all of those because it's like you know i need to like be a more well-rounded normal person um which was fantastic by the way it was so much fun like i remember um for whatever reason um maybe it was just because i was so relatively uneducated to the like the liberal arts but a tale of two cities by dickens at the end like made me bawl my eyes out like crap i i haven't read that it's phenomenal um but yeah so did that you know that took a little bit of time and then i was like okay you know this is all cool but i'm not going to be you know a writer or something like that
and so realized i should make use of this decade of being good at quantitative stuff and either go into um quant finance which is what a lot of mathematicians or physicists do or go into big tech or go into startups that was kind of the the uh decision criteria i actually almost went into quant finance you know in new york but last second had some good advice from some friends which was like you'll always be good at math like don't do it just like stay in your phd and explore um so and prior to then i hadn't taken a formal computer science course i'd learned a lot of you know programming just for physics but ended up taking some classes and i'm very competitive and there was something very competitively satisfying of not having taken cs but going into these grad courses and like doing better than some of the cs students so i kind of kept doing it just for the competitive gratification um and then it wasn't until i took a seminar in program synthesis uh which we now call code generation that just completely nerd sniped me nerd snipe yes what a word i mean there's no better way to put it which is like physics teaches you that things that are fundamental are more important like if it's not fundamental it doesn't
matter like you always want the arbitrary case It's the end dimensional case, you know, the abstract case, that's like where the beauty lies. And there's something very fundamental about, you know, machine learning or AI, not for video or audio or images, which is so arbitrary, right? Like, it's so like, kind of not fundamental, but there's something so deep about machine learning and code for the purpose of creating itself. Like, there's something so just like primitive about that. And that just got me obsessed with that. So stayed at Berkeley for another year to do that. and then it became very apparent that academia is not the place to solve that problem but to solve that problem you kind of need to go into the places where there's very interesting non-trivial code bases and engineering that's happening which is like the industry um and it turns out that the best vehicle to do that is to start a company um and so i was like okay well i don't know anything about starting a company so i bought on amazon again i love going to amazon for books whenever i don't know things clearly but i bought zero to one by peter teal and then i looked up on youtube like how to start a company and like just watched some like random videos one of which had it was
like a video podcast with sean mcguire um and then he was a former physicist and i was like oh i gotta talk to this guy this is your guy and so that's how i ended up sending sean a cold email okay so you reach out to sean mcguire and he responds you guys take a walk yes um at that point like what were you imagining for factory when you were thinking about like what you wanted to build yeah i mean it's it's actually shocking how the initial debt it's like exactly what we initially wanted to do which was like i think at the time it was actually too early but the whole point was our mission from from like the first deck we made the mission statement of the company to bring autonomy to software engineering and so like not ai autocomplete not ai ide's but like full agentic software development to automate the thing and also the thing that i would jokingly say like you know if we're at a bar how would we describe what we do it's like we automate shit that engineers don't want to do yes and it's automate not accelerate and you know ai autocomplete but like just you know dump it to someone else to do um and it's been crazy to see how it's kind of being realized like probably a lot faster than i expected but it is still the same roughly the
same path i think there's so much i didn't know just about enterprise and all hiring and everything
“30 year old cobalt migrations that used to take two years like they can do that in four days with a droid just working on its own”
you can confirm nor deny that it's been crazy this year has been very exciting if i look at ai over the last year and a half it feels like there was this period where ceos are getting a lot more critical on like several aspects which is like a at what point does like the budget not equate to like more productivity totally and b how do we feel about being like super locked in to these labs that are now going up the stack and building all these other features that are starting to be competitive. What have you been hearing from like all these enterprises you're talking to about how they're actually thinking?
And if you were the CEO of a public company, like a public software company, that's not inherently a company, AI company, like, how would you be thinking about this?
“We're now entering phase three, which is like, okay, we have people using AI. Now we need to actually measure the ROI”
“But again, I would take that risk over the risk of like a monopoly to end all monopolies in the United States.”
so the regulation at open ai and anthropic maybe specifically anthropic has been trying to push like what is your what's your take on what the regulation should yes because there's been this funny thing where like there should be some type of regulation but it feels like recently the big labs are all sort of announcing these like crazy dangerous things that the models have been doing and it's almost like my model's so dangerous no my models yeah yeah um which is very toxic by the way this like trying like legitimacy by like how serious the problem is like so quick diatribe i want to get back to policy but a quick diatribe um what happened to a world where if there is like a security incident you are like deeply ashamed and responsible for it not like bragging about it in a research blog post like you should be prosecuted if you have like some of these they made a shitty sandbox and it escaped the sandbox that's your fault that like that's not a thing to brag about like it actually wasn't even it wasn't even that admirable of the agent it was like they gave it a sandbox that was really bad and it escaped the sandbox or there was another instance where they were supposed to not give it access to the internet and say and you know they told
the model we're not giving you access to the internet but then they gave it access to the internet and then it did things on the internet and it's like oh my god it got access it's like this is just like gross negligence and it's somehow being bragged about as if it's like advanced capabilities which these models definitely have advanced capabilities but like we need to separate like negligence of the people who are using and deploying these models and like the models being like actually deeply adversarial yeah um so that's i think one thing that like the narrative right now that is like very troubling because if there is this arms race for how bad it can escalate to like actually people getting hurt or serious economic damage which i don't think is something to be to be bragging about um anyway back to the policy back
“at some point in this decade, there is going to be, like, a billion-dollar incident that occurs because of agents.”
Yes. So we mentioned this, but even at the labs, like basically AI is already writing most of the code, but we're not at 100% yet. Sure.
“It could be the case that we get to RSI and it's just never as good as human assisted like improvement on these things.”
to ask you about companies on a broad sense. Software engineering is becoming way cheaper, way more accessible. People don't need to be technical to have like, you know, like some version of software engineering capabilities now. And then the most technical people are super charged. How do you think this actually translates into the products and services that actually enter our world like what happens to gdp what happens to productivity on all i mean i think
it's it's hard to say because there's like concentric circles that are expanding very rapidly like the broadest one is like the types of problems we can solve um and inside that is like the type of software the average person can build so that one is like what the average person can build is rapidly expanding but then kind of with it the table stakes of what good software is is also growing and also with it is like the scale of problem that we can tackle and i think there are just so many problems that we're not used to thinking of as problems that like like we it's really hard to imagine a world where just like software is like perfect but like There are just so many things. My go-to example is the DMV, especially in California. I've been to California DMV. Recently, they actually redid it, and it's gotten a little bit better.
But eight years ago, it was atrocious. We don't need to live like that. We can have beautiful software everywhere, and it can be in a manner that's completely unnoticeable. like there can just be things where it just like from intention to action it just like works right like and i think it's just so hard to even conceive of some of these things but like just think about any time you might interact with software or hardware and it's like to do like whatever to set up like a bluetooth speaker or to connect the light to this thing or you know send this to that person and you know there are just so many of these things that we can just reduce all this friction and all this bloat and all this like low leverage stuff so that we can stay in like flow state and stay in like high leverage, like do the things that matter.
And do you think everyone's going to build amazing personal software? Like most people still need to have the vision of like, what do they want? Which is actually the hard thing. And it asks the same question on a company level. Like some companies now are like, let's just build it internally. And I've actually talked to like other companies about this who are selling AI and they're like, the sales cycles are actually really long because they're like
“there is nothing that nobody else can build which is crazy but that means the biggest differentiator is like being willing to say no”
because then if you should have done it, it's going to be very loud and clear that you should have. So, yeah. And on a personal level, your journey of building this for the last three years, it started out like you take this walk with Sean he really believes in you hands your check you have like a sequoia check immediately that must have felt pretty hot then there are two years where it must have felt kind of difficult you're like no you have the right vision but product market fit and like things just like aren't quite working yet and then there's this inflection point and things are like you're on a rocket ship what has it felt like to be to have witnessed both sides of that spectrum of like the VC like sun shining on you everyone wants to be your best friend all the way to like you know you're in the thick of like a hard period of the company and I imagine some investors like weren't taking your calls as much and what was that like I mean I think the one the one helpful thing is funny enough going to Princeton people are very much like that there where it's like if there's something they want they'll be very nice
and so that was a nice exposure to like that kind of behavior yeah generally but um but i love princeton but um also there's so many princeton companies now there are yeah like you have to add it to that yeah um but yeah i mean it was it was very difficult i think the important thing is like having people that are fiercely loyal matters so much more than like the fancy names or the fancy logos because at the end of the day like your investors are not going to build the company for you doesn't matter what fancy logo that you have what matters is that you have people that are like deeply in your corner and like when things are difficult sometimes like not sending you like every time a competitor tweets something and instead just being like like you got this like you know let let you cook um yeah it was it was difficult it was it was very convenient also because it just revealed like who are the you know real ones and who's not um and uh yeah i think it's like even today kind of coming out of that is just such a privilege like every like the team that we've assembled we have like over a hundred people these are the smartest people that i've ever worked with who are here to do the work of their lives it is such a privilege and we're in
like the most beautiful city in the damn world like every day it is such a privilege it's so much fun um like i'm just extremely grateful and it's also like we need to work harder like every incremental hour makes such a big difference and thank you for taking this hour but it's really like i actually do appreciate it yeah i swear when i like ask founders that are in these fields right now to like take the hour to the podcast i actually feel it i'm like i hope that this gets out there and people see it and i make the most of their time yeah well a lot some of the smartest people in the world follow you and so hopefully they'll thank you know join factory after hearing this but okay last question on what you just said who are the people that have really um have really shown up as like whether they're on the team or investors so sean mcguire who's on our board i mean he was the very first person who believed in me and is like you know our our first investor and is someone who's just fiercely fiercely loyal um let's see who else comes to mind there's so many great advisors that i've worked with like honestly everyone at sequoia has been fantastic you might think naively like oh they're you know they're the they have the
biggest name so they can be lazy like they earn it every single day they are they grind um Chris Degnan who is the CRO at Snowflake and is an advisor he has taught me like everything about enterprise sales of which I knew nothing going into this um there wasn't a book for that well I that's the thing it's like the books don't do it justice yeah um and what's the biggest lesson you've learned from him. it's like it like i can't even i can't even describe because it's like i literally knew nothing like i did not know a single like there's just so much so many mistakes i made along the way about um sales but i think the important thing is like treat sales as first class or you will not have a legendary business and like it's non-negotiable um nikesh aurora the ceo of palo alto networks has been a great advisor and just like very generous with his time and helping you know guide us Chris Fredrickson who I know you know he's the best he's incredible like he's like so low-key like he's like one of those if you know you know and if you know it's like he is the goat like he's just incredible operator incredible investor and I think he has been someone that like so many people I know too like have relied on and like such a such a deep
way he really is like there to give advice and be there for you oh yeah he's an amazing alex paul from the chain smokers they show up no one everyone at first is like what like they're good investors like no they grind and they are such good investors by the way this is just a hilarious note is like you can see stories from the chain smoker like vegas residency and five years ago it was like hot girls now you looked behind the dj booth and it is literally so clearly just like founders and engineers yeah and i think they're like look people here like don't it it makes people feel good to feel cool and have fun and be around music and like they do offer something different and they're from every founder i've spoken to that works with them they really show up and they like don't sleep like i don't know how like they just like anytime i text like immediate reply like like it makes no sense and we were we were with friends and we ran into someone who was like i've heard about you from ivanka trump she says that you You are the most high-execution founder, and, like, I know that whole team at Affinity loves you, too.
They're incredible. They're also super generous with their time, super generous with their network. Yeah. We're very privileged to have such, like, loyal and, you know, genuine investors. It's, yeah, it's incredible. They all have factory. And Sabrina Halper, of course. One of the OG factory angels. Yeah. Okay, we're going to discuss it. Yeah. Should have been pushier. starting now um but anyway matan thank you so much for coming on the show um it's an honor to have you on and it has been an honor like from the sideline just to see you build this into like such a real and thriving company that is out there in the world droiding away and uh bringing autonomous software engineering to everyone thank you for having me it's a pleasure if you enjoyed this episode please take a minute to subscribe it really helps me grow the show and continue to put out new episodes every week.
