DeepSeek used model distillation to match OpenAI's o1 at lower cost
M.G. Siegler · How DeepSeek Changes AI Research & Silicon Valley w/ M.G. Siegler · Jan 27, 2025 · at 12:00
Investor M.G. Siegler explains how DeepSeek achieved frontier-level AI reasoning performance comparable to OpenAI's o1 model with significantly lower computational resources.
0:00 / 0:28exact quote · 28.9s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“And, you know, they've just used the process of distillation to, you know, effectively bring those bigger versions of the sort of state-of-the-art models and distill them down into you know, smaller models, which eventually led to this R-one, you know, the equivalent of O-one on, on OpenAI's side. And again, for a fraction of the cost, fraction of the compute, and a fraction of the size for these to be able to run.”
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 M.G. Siegler
Sam Altman is positioning OpenAI as a peace broker for Pentagon contracts
“He's trying to position this as like, they are the peace broker here, right? Like that they are the ones who are going to sort of iron out these differences between anthropic and the U S government by cutting their own deal that paves the path. To sort of do a…”
Siegler: Apple lacks the internal mentality and cadence for AI
“One of the reasons that I would worry about Apple going forward is they don't have the right internal mentality about doing this stuff, and you can see it in the sort of products that they've shipped to date, specifically, obviously, with Siri, Where they're c…”
Microsoft Shifted CapEx to OpEx by Extending Data Center Leases
“In this quarter, they decided that they were going to basically change the way that they account for their data center leases. And so historically they've actually Considered them to be 15 years. And that basically meant that they were you know, taking on the …”
Siegler: Michael Burry Shorts Tech Over Server Depreciation Accounting
“Michael Burry you know, famously from Big Short has sort of talked about this as a reason why he's shorting a lot of these companies, because he believes that they're sort of doing funny money, funny math on, on a lot of these a lot of these schedules for Micr…”
Siegler: Frontier AI models face diminishing returns despite massive costs
“Everyone's starting to recognize that there might be a world in which like, you don't want to have the frontier models because you're in charge of the cost of that. And the returns are going to be more and more diminishing over time now.”
Siegler: AI agent landscape will fragment across multiple walled-garden tools
“I worry that we're going to not get a seamless move into like a one single place where we can do all of this, like the sort of the pie in the sky version that you're talking about, where there's one tool, but instead we have to use like Three or four different…”