Amazing. And you're so right. Scaling therapists is hard, especially if your goal is to get care as soon as possible that is high quality, evidence-based to asmany people as possible, as soon as possible.
Organizations don't realize that scaling means not just getting bigger. Scaling is about getting better.It's about scaling excellence.
Now, South Korea is one of America's top allies in Southeast Asia, but now President Trump says he's substantially scaling back joint military exercises with South Korea. Why? Well, because of what he calls his very good relationshipwith the North Korean leader, Kim Jong-un. In a social media post, Mr.
And if we did that more often, I think, parrot less, I think we would be doing way better. Scaling it up means bringing many of these tools into the classroom.And it means rethinking, many times, the academic curriculum.
It will be discussed again at the next board meeting next year, so a democratic ownership of the internet also being part of this discussion. scaling issues, et cetera.
But what ArtCo did was they made it part of their culture. Scaling day was something that people wanted to be a part of because they knew it was part of the future.And so people volunteered.
next one and it's a whole other set of lessons to learn you know especially for me I think the greatest thing is in scaling a business and figuring out how in the world to to do that um every timeI make it to one Milestone I'm kind of like all right I've got everything in my pocket now I'm just going to go and
you're kind of making two points there um the first is uh the idea of of scaling outside of Regulation and I think that unfortunately is sort of the mo of a lot of um a lot of startups sofor example if you've taken on a round of investment someone's breathing down your neck to show returns right and so
So what our task has been with that of the Ministry of Environment and Tourism and the UNDP-- the United Nations Development Program-- Scaling up is what my life is all about.
- And at the time, Jim was associated with the Santa Fe Institute, and he started asking around with, is there anyone up here that's interested in these biological scaling relationships? And then the president at the time said, "You know what? I know of this physicist who's up at Los Alamos who was talking about biological scaling relationships." And so we met Geoffrey and it was like immediately,
finding a line. I think it's about finding a scaling process. - ... the scaling process, but then there is more rapid scaling and there are slower scaling . So innovation, invention, I think is useful to understand so you can predict how likely it is on other planets, for example.
- And of course, now that you have this kind of empirical science/art, you can apply to other more nuanced things like scaling laws applied to interpretability, or scaling laws applied to post-training, or just seeing how does this thing scale . But the big scaling law, I guess the underlying Scaling Hypothesis has to do with big networks, big data leads to intelligence.
for Policing Equity, but our hypothesis is, basically, that we can help do the most good as Google, Google.org, And scaling doesn't mean having to scale your organization.
but when there's a tricky situation. So scaling a business, the hardest thing to do is to keep the ways of working and the values and the principles, keeping them true and alive.
How can we think about using the tools at our disposal to make business more effective, while also creating more meaningful human experiences and scaling those through data and automation? I've had the opportunity to test this idea with a lot of different companies that I've consulted with, spoken with, advised, worked with on different projects over the years.
the scaling of these solutions.
So scaling is important.
The scaling up is very seamless.
And scaling by three orders of magnitude to trillions of pixels per frame to capture all of the historical Landsat, it really
So scaling the small naps and walks out to days and weeks, what's recommended?
And scaling the organization's what's really the big initiative right now.
like scaling the IO and running millions of jobs at the same time.
So scaling smartly, very important for small organizations.
So scaling is not arithmetic, you have to really think calculus, if you will.
You think you're scaling abundance, but we experience it as you scaling scarcity.
- Take a look at this table from Geoffrey West's book. These are the scaling exponents WBE theory predicts, including many that are not multiples of a quarter, but all follow from the same theory. For example, the radius of an animal's aorta should scale with its mass to the three-eighths or .375.
So this one theory accounts not only for the three-quarters power of metabolism, but for literally dozens of other things that biologists have measured. - Some of the scaling laws are easy to explain once you've got the three-quarters law for metabolism. Take a mammal's heart rate, for instance.
So heart rate should scale as metabolic rate over mass. Swapping in the scaling law Kleiber had found, that gives us M to the minus one quarter, meaning bigger animals should have slower heartbeats than smaller ones.
Scaling is about getting better.It's about scaling excellence. You can get big, Mm-hmm.
in order to keep scaling ?
And so the next scaling law is the agentic scaling law. It's kind of like multiplying AI.
where the traditional scaling laws are talked about for pre-training, which is how big your model is and how big your dataset is, and then scaling reinforcement learning,
in the context of scaling laws?
It's all about scaling inference, scaling post-training, scaling context, continual learning, scaling data, synthetic data?
With inference scaling , you don't spend money during training, you spend money later per query, and then it's also like math. How long is my model gonna be
was during inference scaling to achieve peak performance in certain tasks.
of these things, right? Because I keep talking about the scaling , so what is it that's scaling ? What I think is scaling is something I call the cognitive light cone, and the cognitive light cone is the size of the biggest goal state that you can pursue. This doesn't mean how far do your senses reach? This doesn't mean how far can you affect it? So the James Webb
- Do you think the scaling laws are holding strong,
So I do think scaling laws are working, but it's tough to get, at any given time, the models we all use the most
- This is a scaling law shirt, by the way.
is going to continue and that there's some magic to it that we haven't really explained on a theoretical basis yet. - And of course the scaling here is bigger networks, bigger data, bigger compute. - Yes. - All of those.
like scaling laws applied to interpretability, or scaling laws applied to post-training, or just seeing how does this thing scale . But the big scaling law, I guess the underlying Scaling Hypothesis has to do with big networks, big data leads to intelligence. - Yeah, we've documented scaling laws in lots of domains other than language, right?
we've been able to look inside these systems and understand what we see, right? Unlike with the scaling laws where it feels like there's some, you know, law that's driving these models to perform better, on the inside, the models aren't, you know, there's no reason why they should be designed for us to understand them, right?
talk about scaling , I I will just say we have a plan. We have what it takes
But responsible scaling isn’t the only precaution labs can take.
Enter AI benchmarks and scaling laws.
We call those formulas scaling laws, and so far, they help us predict what future AI with even more data and more compute might be able to do.
It definitely is scaling because we have a fintech platform.
wisdom around leadership and scaling companies.
And what kinds of scaling issues did you face?