Machine Learning

“And to do that, you have to underwrite. Like, you can't shortcut the, I'll just look at your FICO score, or I'll just sort of, you know, I'll look at your Facebook friends. Like, none of that works. You actually have to do a real, very sophisticated degree of …”

“AIs seem Significantly better at engineering and grungy stuff and sort of just keeping trying than they seem to be at conceptual breakthroughs, but their ability to do sort of these. Breakthroughs, especially in easy to verify domains are improving. And like, …”

“The place that everybody was coming to it from was just assuming that we'll never have a lot of data for robotics, and we need to do everything we can to try and design our system smarter, or like, figure out tricks along the way to try and get robotics starte…”

“You need to have data to learn stuff, and everybody's robots were just sitting still, and it just felt like we have to get on this path where like, we're actually moving and physically interacting with the world at scale, and then we'll figure out all the rest…”

“People think you can't work on like AI bio unless you're a biologist, but it's kind of like you can't work on like video models unless you're like a director or something. Like there are all these like super domain specific things like, oh yeah, to understand …”

“I mean, I think the example you raised of a thousand X difference in bringing moving data versus actually computing on it in, in terms of energy is, is a pretty significant one. And it shapes a lot of aspects of what we do in machine learning. Because if you d…”

“And ultimately, like less than five percent of our customers had enough data to really like for a machine learning model to even be useful to them.”

“We've built a AI tech enabled platform that creates outcome intelligence. And so activate becomes not just that one day and then 100 leaders. We now have some permanence because we've provided some actual data of the cities that we've had these ideations in cr…”

“One hot take that I have is that I've noticed a lot of companies, like, kind of over-engineering their, like, ML usage. Like, they'll have, like, They'll build their own custom models. They'll try to train their own models when really, like, you should, you co…”

“I think machine learning tends to have a long history of people starting with bio-plausible arguments, and then realizing that there's some variant of them that seems highly bio-implausible that actually works better.”

“Well, in 22 at the very bottom, we said, we're on an older version of machine learning. We're going to completely throw out our technology, rebuild it, and go to what is really cutting edge and current and in the field of recommendation systems.”

“As long as the capabilities frontier keep moving and we want a healthy, independent ecosystem, we'll just keep Raising more capital. There's no end to that. I don't really, the day machine learning stops working as a systematic way to give humanity more capabi…”

“I think the major reason this product became so successful had to do with machine learning infrastructure that was completely under the surface and really hard to see.”

“And I think the physical sciences, physical engineering, Will have a very similar property where we establish these scaling properties and Bring that mindset.”

“And I think increasingly what we'll see is a shift where domains where collecting data is straightforward They actually end up falling into the easy bucket over time, even if they are physically intricate. But there will be domains where collecting data is dif…”

“I personally don't think that machine learning or AI in 50 years is still going to be built on this stack.”

“Half the problem in any of these products is actually not the machine learning stuff. It's actually just the data, like Getting, getting it, being able to work with it at scale, being able to apply it, being able to do that in the context of like a real runnin…”

“I think one of the areas where machine learning kind of just with what's out there right now has the most promising chemical sciences is in solving multi-dimensional challenges.”

“Most people who actually work on Getting materials to the device scale, say something that would be in your television or something like that, is they will tell you that it's not just the material, it's the process. And I think we're at ground zero. We're nowh…”

“Some research sub-fields are trying to do that, but it's not really developed across material science.”

“Historically in machine learning, you always, you know, it's like the rule was you have to know your data set really well. But now we're kind of outsourcing that to the AI itself, where the AI is the, it's the AI's job to understand the dataset and figure out …”

“I mean, the last wave of AI was machine learning, 1015 years ago now, 10 years ago, really. That was not transformative across everything. It was a new bunch of stuff that everybody could build, and there were a bunch of new companies with it, but it wasn't th…”

“That's where it gets harder. You know, when you're chasing from 99% accuracy to 99.9. Gen AI still doesn't fully help with that. So you still have to kind of do more typical machine learning”

“This theory called maximum entropy. And it's about like, how do you take complex simulations and match them to limited observations? And it's like the inverse of machine learning. Machine learning is like, you have simple models, you're going to a lot of data …”

“That means you have a giant black box and the giant black box nature of it is the reason that they hallucinate, make things up and behave in unpredictable ways.”

“Generally, I'm not really, I don't really believe that model thoughts have to be the same with human thoughts. I'm actually like, generally in ML, I'm more of the school of thought of let the model do whatever it wants.”

“ML. ML can be learned easily. Our knowledge can be learned easily.”

“So what's happening now is we're starting to realize that the Gen AI adoption paradigm in the enterprise works the same way that ML does.”

“I think we're still having trouble not at OpenAI, but I think as a whole, producing lots of people that do lot, want to do lots of both systems work and ML work. And I think if you're trying to push the frontier, you don't know which Place is currently bottlen…”

“At MDT, we've been using these machine learning tools since 2001. So we have a 24 year head start on someone who is new to the game.”

“Machine learning works best on the training set distribution. But in science and technology, we almost only care about auto-domain generalization, right?”

“I used to be quite skeptical that neural networks were necessarily the ultimate A solution to machine learning. I seem to be quite wrong on this one.”

“We don't actually use the large language models today in our product. So this is more traditional machine learning, you know, from proprietary data like time series data.”

“One of the weak signals that he intercepted years ago was a wonky but exciting development in machine learning that kept popping up in the emails. Jensen decided that Nvidia needed to invest more in tools for accelerating workloads on its GPUs.”

“On the material science side, we're seeing scaling laws within physics, within chemistry both with respect to simulations, with respect to experiment, and it's like the same kind of principles at play and ML.”

“In the precision system, one of the inherent advantages is that the data is so regular in structure that we're able to compress it. We're able to learn across patients, across populations, and leverage those learnings in the machine learning algorithms that we…”

“For product and engineering and design people and, you know, those kinds of functions, I actually think that if you are just curious about the stuff works, it doesn't matter at all if you've never done it before. In fact, if you were to filter for people who'v…”

“On the government side specifically, I think there's been a lot of like false goods that have been sold, predictive maintenance, prognostic, you know, type things that when you actually get down to it, like the data, the telemetry and the sensor data and the l…”

“With machine learning, what you're able to do is reduce costs by 99% and increase resolution by a hundred times.”

“With machine learning, what you're able to do is reduce costs by 99% and increase resolution by a hundred times.”

“Like machine learning is the stuff that's now become easy and then AI is all the fun new stuff. And then as soon as it stops becoming the cutting edge, Then it just becomes, oh, that's just machine learning.”

“And it was a somewhat localized problem because like you're doing data science on your laptop versus now I'm just shipping off secret IP to some SaaS company or something like that.”

“These observations were made by skilled geologists, and the rocks haven't moved, so there's no expiration date on the data, and so you can take data sets like this that provide ground truth and use it for training machine learning models based on modern airbor…”

“There are so many other ways to do machine learning that are like, you know, support vector machines, linear regression, logistic regression, there's like a whole bunch of techniques, but it happens to be that neural networks is the one way to do things when y…”

“By the end of the year, Right? Most code is going to either be refactored or written from the start using machine learning. It's going to be coding agents that are helping to do that.”

“Typically in machine learning, a lot of the successful applications of like recommender systems, language models like image detection, a lot of the consumers of that Of the model outputs are actually humans who could actually check it, and the humans are good …”

“But really, I think with some of these more product experience questions, there's something qualitative about it that is very hard to quantify. That is one of the big challenges internally, actually, is how do you hill climb effectively on what is really an ML…”

“I think from my perspective, good taste in ML today, because it's such a fast moving field with so many people working across, you know, open source and APIs and big labs and so forth. Really, you're trying to identify What part of the ecosystem or what part o…”

“And also, you know, another, another BS thing that Apple pitched, which this is privacy thing, which honestly, like, I cannot believe how much mileage they got out of that. They kind of just painted themselves into a corner because, you know, all these other c…”

“I think that Those are the buzzwords today that, like, blockchain was a year ago, and that big data was a year ago.”

“The vast majority of, you know, mechanistic data that's been generated to date is really made to ask very specific, very well scoped questions and just, you know, way more tokens per experiment is, you know, just, it's just going to be the way to do it.”

“I transitioned to product design, which is a rare transition, I think. But it actually, like, really helped me in my path to starting this company. Because think about it, product design plus machine learning, like, hard to beat as a combination.”

“If you have structured data, you're doing forecasting models, a lot of the fraud management, traditional artificial intelligence, machine learning, Is going to be more efficient, more effective, and certainly more cost effective as a way to do it.”

“There's a business analogy of this, which is that, or not actually an analogy, it's called Goodhart's law. It's literally the same thing as overfitting, just for businesses. Goodhart's law just says any metric that becomes a goal ceases to be a good metric.”

“With machine learning, first of all, you see that the performance of the models follows an S-curve. So it's not like it just goes off to infinity, right? And the S curve, it kind of plateaus around human level performance.”

“There was almost nothing that went superhuman except for something like AlphaGo.”

“It's a vast majority of GF-AI systems have seen, like, you have, like, traditional or, like, analytical ML components with GF-AI.”

“I definitely see a lot of people building very good applications without traditional ML background.”

“One of the weak signals that he intercepted years ago was a wonky, but exciting development in machine learning that kept popping up in T five T emails. Jensen decided that NVIDIA needed to invest more in tools for accelerating workloads on its GP use. This is…”

“If you don't have measurements, like in the old machine learning, whatever metrics you use, you're not going to advance. You're going to have V-one and then you're going to have V-two, and you have no way to know if you've made a progress.”