Scott: AI scaling laws have not hit a limit yet
Kevin Scott · Kevin Scott, CTO @ Microsoft: An Evaluation of Deepseek and How We Underestimate the Chinese · 20VC with Harry Stebbings · Mar 31, 2025 · at 8:31
Kevin Scott is the CTO of Microsoft. He discusses whether AI model development is reaching a performance ceiling or diminishing returns regarding scaling laws.
0:00 / 1:29exact quote · 89.2s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“I can very clearly see what we're doing now and like what we're doing next. And I don't see the limit to the scaling laws. Like if you're just sort of thinking about the raw capability of the models and like how well you can condition them to reason over increasingly complicated things like, and I'm sure like There at some point will be maybe a limit. Like I intuitively feel like there must be, there are some people who don't believe that there's a limit that, you know, if you, the limit that human beings have on intelligence is like, you've got so many neurons like packed into your skull and like, you've got about a 20 watt power envelope and like, that's the limit. Some people believe that, you know, if you're You have AIs that, like, there is no such limit and, like, things, you know, will continue to scale into, like, you know, weird territory. I don't really, like, that I don't necessarily believe. I believe we will get to some point where we'll hit a scaling asymptote and, you know, like, there'd just be diminishing marginal returns and, like, it's so expensive that we will decide it's not worth spending that next dollar to, like, make Make this thing one unit smarter because we haven't, you know, figured out how that translates into something that's useful for like the people who are using the tool. I think that point will come. I don't, I just don't see it yet. Like it's not. It's not, not in the viewfinder right now.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Kevin Scott
Scott: Microsoft chose not to launch models more interesting than DeepSeek-R1
“We've had models more interesting than Deep Seek R-One that, like, we didn't even We chose not to even launch them.”
Scott: 95% of net new code will be AI-generated within five years
“95% is going to be AI generated. I think very little is going to be line by line is going to be human written code.”
Scott: Frontier AI models diagnose health issues better than average GPs
“It is already the case that I think the frontier models are probably better like health diagnosticians than Your average GP is”
Scott: AI's economic value will accrue to products, not infrastructure
“But most of the value has to be in the products. Like, you know, we don't build infrastructure just for the sake of infrastructure. Like we build infrastructure so people can make product.”
Scott: Claims about proprietary data's value for AI lack scientific backing
“Most of those assertions that people make are like most of the assertions that people make are just unfounded in any kind of science. Like no, no one's gotta, and like the measurements we do have show that there's a pretty big disconnect around what some peopl…”