Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

SITUATION EXPLAINED: How much are frontier labs actually spending on training data? .Sean Cai: "Frontier labs are spending about $10 to $15 billion per lab on data." "Really good long horizon tasks go up to $20,000 each. A complete browser-use version of SAP was rumored at $500,000." "Despite everybody... show more

MTS

165,952 subscribers

334,215 views • 28 days ago •via X (Twitter)

Science & Technology Finance

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Yesterday I interviewed Sean Cai about AI data. This is essentially a guide for founders on how to sell data and RL envs to AI labs. "I've never seen a data contract get turned down by a top lab, if it's good quality data, for budget reasons." 00:00 What areas of data are underserved? 02:10 For bio data, is it real-world or purely digital? 04:21 For cyber data, which subsets are most underserved? 05:50 What is the sales process like? 07:04 Why would a lab not renew or increase their purchase volume? 10:13 When a researcher is exploring a new direction, what's the first step? 11:35 In robotics data, what do you view as underserved? 13:12 What does the initial data delivery look like, what format? 13:53 Do labs have more sophisticated internal setups for running environments? 14:32 Are the non-frontier labs buying off-the-shelf data from Anthropic / OpenAI vendors? 16:11 Do Anthropic data vendors put expiry timeframes on the exclusivity? 16:42 Are purchase decisions researcher-led? 17:41 Decagon, Sierra, Ramp: what kinds of data are they buying? 19:06 Long-term, when do labs still need to buy external data vs train on user traces? 21:15 Will end-vendor benchmarks shift to performance per dollar? 22:04 How many labs are spending at the 1B+/yr data level? 23:53 Delta between Anthropic's stated $1B and your 10-20B/lab number? 26:05 What makes inference providers / neoclouds a good fit to acquire RL env cos?

Yesterday I interviewed Sean Cai about AI data. This is essentially a guide for founders on how to sell data and RL envs to AI labs. "I've never seen a data contract get turned down by a top lab, if it's good quality data, for budget reasons." 00:00 What areas of data are underserved? 02:10 For bio data, is it real-world or purely digital? 04:21 For cyber data, which subsets are most underserved? 05:50 What is the sales process like? 07:04 Why would a lab not renew or increase their purchase volume? 10:13 When a researcher is exploring a new direction, what's the first step? 11:35 In robotics data, what do you view as underserved? 13:12 What does the initial data delivery look like, what format? 13:53 Do labs have more sophisticated internal setups for running environments? 14:32 Are the non-frontier labs buying off-the-shelf data from Anthropic / OpenAI vendors? 16:11 Do Anthropic data vendors put expiry timeframes on the exclusivity? 16:42 Are purchase decisions researcher-led? 17:41 Decagon, Sierra, Ramp: what kinds of data are they buying? 19:06 Long-term, when do labs still need to buy external data vs train on user traces? 21:15 Will end-vendor benchmarks shift to performance per dollar? 22:04 How many labs are spending at the 1B+/yr data level? 23:53 Delta between Anthropic's stated $1B and your 10-20B/lab number? 26:05 What makes inference providers / neoclouds a good fit to acquire RL env cos?

Chris Barber

128,976 views • 1 month ago

In 2024, Stanford's Yann Dubois gave a 104-min masterclass on how frontier LLMs are actually built. 99% of people still don't know this. His frameworks: - Data beats model cleverness - Perplexity isn't product quality - Alignment changes behavior, not IQ 12 insights: 🧵

In 2024, Stanford's Yann Dubois gave a 104-min masterclass on how frontier LLMs are actually built. 99% of people still don't know this. His frameworks: - Data beats model cleverness - Perplexity isn't product quality - Alignment changes behavior, not IQ 12 insights: 🧵

andrew engler

66,856 views • 2 months ago

Introducing Contra Labs. The first frontier data and evaluation lab for Creative AI.

Introducing Contra Labs. The first frontier data and evaluation lab for Creative AI.

ben

445,995 views • 2 months ago

Andrej Karpathy had already talked about this possibility 1 month back, that he might join one of the big AI labs and why he may do so. "They (the AI labs) are kind of like at the edge of capability of what's possible. And they are working on what's coming down the line. And I think if you're outside of that frontier lab, your judgment fundamentally will start to drift because you're not part of what's coming down the line. And so I feel like my judgment will inevitably start to drift as well. And I won't actually have an understanding of how these systems actually work under the hood. It's a big system. I won't have a good understanding of how to develop and etc. And so I do think that in that sense, I agree and something I'm nervous about. I think it's worth basically being in touch with what's actually happening and actually being in the frontier lab. And if some of the frontier labs would have me come for some amount of time and do really good work for them, then I think that's maybe a good setup" --- From 'No Priors: AI, Machine Learning, Tech, & Startups" YT channel (link in comment)

Andrej Karpathy had already talked about this possibility 1 month back, that he might join one of the big AI labs and why he may do so. "They (the AI labs) are kind of like at the edge of capability of what's possible. And they are working on what's coming down the line. And I think if you're outside of that frontier lab, your judgment fundamentally will start to drift because you're not part of what's coming down the line. And so I feel like my judgment will inevitably start to drift as well. And I won't actually have an understanding of how these systems actually work under the hood. It's a big system. I won't have a good understanding of how to develop and etc. And so I do think that in that sense, I agree and something I'm nervous about. I think it's worth basically being in touch with what's actually happening and actually being in the frontier lab. And if some of the frontier labs would have me come for some amount of time and do really good work for them, then I think that's maybe a good setup" --- From 'No Priors: AI, Machine Learning, Tech, & Startups" YT channel (link in comment)

Rohan Paul

54,762 views • 1 month ago

I gave a talk at ICLR 2026 about how we are scaling RL on frontier LLMs with 1T+ parameters, on experimental data from our physical lab at Periodic! Here's a rough recording of the talk:

I gave a talk at ICLR 2026 about how we are scaling RL on frontier LLMs with 1T+ parameters, on experimental data from our physical lab at Periodic! Here's a rough recording of the talk:

Rishabh Agarwal

218,345 views • 1 month ago

Traverse (Traverse (YC W26)) is a data research lab solving the hardest problem in AI training: subjective, taste-dependent work. They generate long-horizon training data for frontier models by observing and capturing what specialists actually do on the job. Congrats on the launch, Lance Yan and Zachary Yu!

Traverse (Traverse (YC W26)) is a data research lab solving the hardest problem in AI training: subjective, taste-dependent work. They generate long-horizon training data for frontier models by observing and capturing what specialists actually do on the job. Congrats on the launch, Lance Yan and Zachary Yu!

Y Combinator

75,533 views • 3 months ago

tom stallard on mclaren sharing all data between both drivers: “all of the data is available to both cars within mclaren and actually within our race engineering group we have a really good, and across the whole team as well [..] an incredibly collaborative relationship and it's very much like how do we use the data to make the team go faster. so, although some of it we're using to coach the drivers, i've not seen any kind of efforts from either driver to kind of hide things from their teammate or anything like that. actually, we very much use that to just push the next level up. so, as much as they want to beat each other, i think they both have quite good awareness that – like oscar wants to be the best version of lando”

tom stallard on mclaren sharing all data between both drivers: “all of the data is available to both cars within mclaren and actually within our race engineering group we have a really good, and across the whole team as well [..] an incredibly collaborative relationship and it's very much like how do we use the data to make the team go faster. so, although some of it we're using to coach the drivers, i've not seen any kind of efforts from either driver to kind of hide things from their teammate or anything like that. actually, we very much use that to just push the next level up. so, as much as they want to beat each other, i think they both have quite good awareness that – like oscar wants to be the best version of lando”

ray

169,554 views • 7 months ago

Lori Chavez-DeRemer on jobs data: "Sometimes it's too late to get accurate data points ... we don't have that accurate data for the Fed to make a good decision, but I believe the Fed needs to cut interest rates again in December to really prop up this economy."

Lori Chavez-DeRemer on jobs data: "Sometimes it's too late to get accurate data points ... we don't have that accurate data for the Fed to make a good decision, but I believe the Fed needs to cut interest rates again in December to really prop up this economy."

Aaron Rupar

158,278 views • 7 months ago

Are Data Centers Worth the Noise to Local Communities? "Multiple cities of people have filed lawsuits and complaints against these data centers because of the noise." "The first question I ask myself is, are we gonna need data centers in America? Yes, if we wanna compete in the modern world, in the modern economy, we are going to need them. They are good for society and good for the economy."

Are Data Centers Worth the Noise to Local Communities? "Multiple cities of people have filed lawsuits and complaints against these data centers because of the noise." "The first question I ask myself is, are we gonna need data centers in America? Yes, if we wanna compete in the modern world, in the modern economy, we are going to need them. They are good for society and good for the economy."

PBD Podcast

32,573 views • 1 month ago

Here is the full run of MicroFactory, that autonomously assembles a photo frame - a real product sold on Amazon. This is a good example of how to deal with complex tasks, while we do not yet have a large set of robotic data to train a big general ai model. Thread:

Here is the full run of MicroFactory, that autonomously assembles a photo frame - a real product sold on Amazon. This is a good example of how to deal with complex tasks, while we do not yet have a large set of robotic data to train a big general ai model. Thread:

Igor Kulakov

197,003 views • 7 months ago

What if we stopped calling them data centers? Samo Burja: "Consider why do we talk about data centers, not supercomputers?" "A data center in my hometown that sounds like someone's gonna steal my data. I don't feel like I have any opportunities." "A supercomputing center in my hometown? Suddenly I might have opportunities. I might have a job. Galaxy-brain stuff is happening there." "These are not data centers anymore, these are actually distributed supercomputers." "If the median use case is not the laptop on my desk or the phone in my hand, the chips can be pretty big. You might actually see the macro chip. I think people are underrating that future and not really expecting it."

What if we stopped calling them data centers? Samo Burja: "Consider why do we talk about data centers, not supercomputers?" "A data center in my hometown that sounds like someone's gonna steal my data. I don't feel like I have any opportunities." "A supercomputing center in my hometown? Suddenly I might have opportunities. I might have a job. Galaxy-brain stuff is happening there." "These are not data centers anymore, these are actually distributed supercomputers." "If the median use case is not the laptop on my desk or the phone in my hand, the chips can be pretty big. You might actually see the macro chip. I think people are underrating that future and not really expecting it."

MTS

12,637 views • 22 days ago

What could be the next frontier in data analytics for rugby? How about "off the ball" data? Here is a sample of an animation generated from player GPS data - this was done to investigate defensive connectivity originating from a lineout.

What could be the next frontier in data analytics for rugby? How about "off the ball" data? Here is a sample of an animation generated from player GPS data - this was done to investigate defensive connectivity originating from a lineout.

Simon Chi

124,447 views • 2 years ago

Over the last few months, we’ve been thinking about how to learn from “off-domain” data - data from non-robot sources like video or simulation. These data sources are not quite good enough to learn policies (even monolithic VLA models) directly, but they still contain lots of information that can be useful for generalizable robot control. How can we develop robot learning models that are able to make use of this type of data for generalizable control? In new work, that we call HAMSTER, we show that VLMs can be useful for enabling robotic learning from off-domain data, but specifically when used through hierarchical VLA architectures. We show that this class of models can learn generalizable robot policies for the real world from large-scale, off-domain data. A 🧵 (1/10)

Over the last few months, we’ve been thinking about how to learn from “off-domain” data - data from non-robot sources like video or simulation. These data sources are not quite good enough to learn policies (even monolithic VLA models) directly, but they still contain lots of information that can be useful for generalizable robot control. How can we develop robot learning models that are able to make use of this type of data for generalizable control? In new work, that we call HAMSTER, we show that VLMs can be useful for enabling robotic learning from off-domain data, but specifically when used through hierarchical VLA architectures. We show that this class of models can learn generalizable robot policies for the real world from large-scale, off-domain data. A 🧵 (1/10)

Abhishek Gupta

11,994 views • 1 year ago

Powell: The point is that we are starting to see layoffs and important gatherers of data are saying that they are having to cut back on the size of their surveys and that will lead to more volatility in the surveys. I think we should take a step back. From our standpoint, having really good data on the state of the economy at any given time is a huge public good.

Powell: The point is that we are starting to see layoffs and important gatherers of data are saying that they are having to cut back on the size of their surveys and that will lead to more volatility in the surveys. I think we should take a step back. From our standpoint, having really good data on the state of the economy at any given time is a huge public good.

Acyn

32,876 views • 1 year ago

On the heels of a weak August jobs report that fell well short of economists’ expectations, White House economic adviser Kevin Hassett is downplaying the jobs data, telling Margaret Brennan “the reason we are getting massive revisions and numbers that don't make sense is really that we have got to modernize the way we do the labor data.” Hassett goes on to add, “The thing I am most suspicious of in terms of data quality is the job number, in part, because there are other ways to do it.”

On the heels of a weak August jobs report that fell well short of economists’ expectations, White House economic adviser Kevin Hassett is downplaying the jobs data, telling Margaret Brennan “the reason we are getting massive revisions and numbers that don't make sense is really that we have got to modernize the way we do the labor data.” Hassett goes on to add, “The thing I am most suspicious of in terms of data quality is the job number, in part, because there are other ways to do it.”

Face The Nation

17,984 views • 9 months ago

Author of Designing Data-Intensive Applications, Martin Kleppmann, on why engineers don't get to ignore ethics just because the work is technical — a theme he felt strongly enough about to address in both editions: “I felt that ethics had been quite ignored as a concern during my time in industry. Especially in startups, people were very focused on building a product that their customers would love and deprioritized ethical questions in the process. For example, with consumer facing products, it might be that the products are very much geared towards essentially data harvesting, collecting behavioural data, because that's what can be monetized in the form of advertising. There seemed to be very little reflection on what was good and bad about these sorts of things. I really wanted to encourage a bit of thinking there. I didn’t want to prescribe a particular approach, but to point out there is this thing such as data protection legislation now, which we do have to think about in the architecture of our data systems and there is an ethical responsibility.”

Author of Designing Data-Intensive Applications, Martin Kleppmann, on why engineers don't get to ignore ethics just because the work is technical — a theme he felt strongly enough about to address in both editions: “I felt that ethics had been quite ignored as a concern during my time in industry. Especially in startups, people were very focused on building a product that their customers would love and deprioritized ethical questions in the process. For example, with consumer facing products, it might be that the products are very much geared towards essentially data harvesting, collecting behavioural data, because that's what can be monetized in the form of advertising. There seemed to be very little reflection on what was good and bad about these sorts of things. I really wanted to encourage a bit of thinking there. I didn’t want to prescribe a particular approach, but to point out there is this thing such as data protection legislation now, which we do have to think about in the architecture of our data systems and there is an ethical responsibility.”

The Pragmatic Engineer

783,831 views • 1 month ago

Dr. Marty Makary is Calling for More Data Before Approving More COVID Shots “Should a 12-year-old, thin, healthy girl get a seventh Covid shot this year? I don't know the answer to that because we don't have good clinical data. We can get that data.”

Dr. Marty Makary is Calling for More Data Before Approving More COVID Shots “Should a 12-year-old, thin, healthy girl get a seventh Covid shot this year? I don't know the answer to that because we don't have good clinical data. We can get that data.”

Chief Nerd

160,084 views • 1 year ago

How do we cost-plus data for millions of Americans? Noble Mobile was inspired in large part by Mark Cuban’s Mark Cuban Cost Plus Drug Company - it turns out Americans are spending twice as much on our wireless data as people in other countries.

How do we cost-plus data for millions of Americans? Noble Mobile was inspired in large part by Mark Cuban’s Mark Cuban Cost Plus Drug Company - it turns out Americans are spending twice as much on our wireless data as people in other countries.

Andrew Yang🧢⬆️🇺🇸

35,435 views • 9 months ago

Dr. Larry Palevsky: “Vaccines are MADE to cause inflammation.” “Every child vaccinated is injured because inflammation has to happen.” “By definition, we cause inflammation when we give vaccines.” “How long does that inflammation go on?” “We have no data.” “Where does that inflammation go in the body?” “We have no data.” “When you look at the studies… there are no real understandings of what happens.” “We have so many kids with chronic inflammation.” “We are administering a medical procedure that causes inflammation without data to understand how long it lasts and where it goes in the body.” “We owe it to our children to find these answers before we keep causing inflammation.” MAHA Institute

Dr. Larry Palevsky: “Vaccines are MADE to cause inflammation.” “Every child vaccinated is injured because inflammation has to happen.” “By definition, we cause inflammation when we give vaccines.” “How long does that inflammation go on?” “We have no data.” “Where does that inflammation go in the body?” “We have no data.” “When you look at the studies… there are no real understandings of what happens.” “We have so many kids with chronic inflammation.” “We are administering a medical procedure that causes inflammation without data to understand how long it lasts and where it goes in the body.” “We owe it to our children to find these answers before we keep causing inflammation.” MAHA Institute

Children’s Health Defense

74,609 views • 3 months ago

For too long, advanced data optimization was reserved for frontier labs. Now it’s for builders everywhere. Control. Shape. Own.

For too long, advanced data optimization was reserved for frontier labs. Now it’s for builders everywhere. Control. Shape. Own.

adaption

48,719 views • 3 months ago