Transcription: OpenAgents Episode 181 - American DeepSeek?

Source: https://x.com/OpenAgentsInc/status/1941327986327740800 Wiki source: https://raw.githubusercontent.com/wiki/OpenAgentsInc/openagents/Video-Series.md Media title: OpenAgents - @MistralAI Episode 181: American DeepSeek? We read & review "The... Upload date: 20250705 Transcription model: gpt-4o-transcribe-diarize Generated at: 2026-06-01T04:05:34Z

Machine-generated transcript. Review speaker labels and wording before using this as quote-grade source material.

[00:00] Christopher David: Happy Independence Day.

[00:01] Christopher David: I'd like to go through this article,

[00:03] Christopher David: the American Deep Seek Project that just came out today.

[00:07] Christopher David: A little bit of background,

[00:08] Christopher David: this is an idea that's been floating around for a while.

[00:11] Christopher David: An article came out in April called How to Build an American Deep Seek and Why We Need It Now.

[00:16] Christopher David: It's a good read.

[00:17] Christopher David: And I agreed with all of this and I didn't feel any particular urgency to do anything specific for this in part because it was like, oh.

[00:25] Christopher David: You know,

[00:26] Christopher David: Meta is kind of this big institutional anchor.

[00:29] Christopher David: They're going to be pouring all their money into building these open models and the community will kind of take that llama release and spin it and fine tune it and we'll be able to collectively catch up to the big labs.

[00:41] Christopher David: Look at this prediction.

[00:43] Christopher David: Llama 4 will likely deliver amazing results.

[00:45] Christopher David: It didn't.

[00:47] Christopher David: It didn't. It was so bad.

[00:51] Christopher David: The meta is now reconsidering their open approach to AI.

[00:56] Christopher David: This came out last week.

[00:59] Christopher David: Executives in meta's AI division discussed de-investing in its AI model,

[01:04] Christopher David: LAMA,

[01:04] Christopher David: two people familiar with the discussion said.

[01:07] Christopher David: They discussed embracing AI models from competitors like OpenAI and Anthropic.

[01:17] Christopher David: What other signs do you have like their spokesperson was like no no we're committed to it Yeah, they don't want a huge flight of people away from Lama

[01:24] Christopher David: But is the writing on the wall well they're in the process of trying to hire these

[01:30] Christopher David: two gentlemen who are the largest funders of closed AI labs outside of the big labs

[01:38] Christopher David: Mentioned them in our episode 124 one of the companies they funded

[01:44] Christopher David: Raised $465 million,

[01:46] Christopher David: shipped nothing but a blog post a year ago,

[01:49] Christopher David: have done absolutely nothing except for call for regulation.

[01:56] Christopher David: As well as SSI, who has no customers other than the government.

[02:01] Christopher David: And that closed AI guy who was talking about nationalizing the AI lab. So that's who Zuck is trying to poach.

[02:10] Christopher David: Gross just stepped down from SSI.

[02:12] Christopher David: He's probably going to be joining meta.

[02:13] Christopher David: Okay, this is hugely problematic.

[02:20] Christopher David: It's safe to assume that meta is now just a bad actor.

[02:23] Christopher David: They're leaving team open.

[02:27] Christopher David: I'm happy to be surprised if they come out with some super strong statement or new model,

[02:31] Christopher David: but it looks like they're ditching team open.

[02:34] Christopher David: So, I'm now feeling the urgency.

[02:37] Christopher David: Here, let's go through this article from Nathan Lambert,

[02:40] Christopher David: what

[02:43] Christopher David: I think the next goal for the open source AI community is.

[02:47] Christopher David: While America has the best AI models in Gemini, CloudO3, etc., and the best infrastructure with NVIDIA, it's rapidly losing its influence over the future directions of AI that unfold in the open source and academic communities.

[02:59] Christopher David: Chinese organizations are releasing the most notable open models and data sets across all modalities,

[03:03] Christopher David: from text to robotics with video,

[03:05] Christopher David: and at the same time it's common for researchers worldwide to read far more new research papers from Chinese organizations rather than their Western counterparts.

[03:13] Christopher David: Um,

[03:15] Christopher David: so I'm not going to read all this.

[03:18] Christopher David: I do want to go down to some stuff at the bottom.

[03:21] Christopher David: But, um, how's this doing?

[03:23] Christopher David: Can I show you this?

[03:25] Christopher David: Don't look at this. Don't look at this.

[03:28] Christopher David: But I have been enjoying Deep Seek R1.

[03:31] Christopher David: I love Deep Seek.

[03:33] Christopher David: I love this Chinese lab Deep Seek. Their model is awesome.

[03:40] Christopher David: Say hello,

[03:41] Christopher David: deep seek.

[03:46] Christopher David: This is a little work in progress.

[03:49] Christopher David: Anyways, we are integrating deep seek alongside,

[03:52] Christopher David: you know, U.S. models and going to be doing sort of a best of all worlds.

[03:57] Christopher David: I'll do a separate announcement about that soon.

[04:01] Christopher David: So we need an open source version of deep seek. And I want to kind of jump down to some stuff here about why I'm...

[04:09] Christopher David: really bullish about this approach generally in part because open agents I think are going to help be that sort of like magic x-factor that helps to make up for the lack of the insane amount of resources in training based models.

[04:26] Christopher David: Here we go.

[04:27] Christopher David: The efficiencies of open source software style development are dramatically stronger for agentic systems than models.

[04:34] Christopher David: Models are singular entities built with expensive resources and incredible focus.

[04:40] Christopher David: Agents are systems that can use many models off the shelf and route requests depending on what's needed.

[04:46] Christopher David: Speaking of routing,

[04:47] Christopher David: I literally had my agents today build a router using EffectTS,

[04:56] Christopher David: routing between different models and agents, all in Effect.

[05:02] Christopher David: Uh we'll open source this pretty soon. Uh but anyways, they're speaking my language here.

[05:09] Christopher David: The agentic era is the opportunity open models have needed, but we need to clear much stronger performance thresholds before the open counterparts are viable.

[05:18] Christopher David: I don't even know how much of this

[05:20] Christopher David: requires model level innovation um what's out there is already like quite good but hey i'll appreciate further compute at the base layer uh we have companies like openai and google launching cloud code competitors that pretty much flop yep yep yeah

[05:39] Christopher David: we're doing our own you have to be neutral between models you can't just like push your own models maybe anthropic can anyways anyways

[05:47] Christopher David: For this reason we have finite time to get there.

[05:57] Christopher David: The best American models are plagued by the llama license and rumors that future versions will be discontinued. Yeah, this is done.

[06:05] Christopher David: The llama ecosystem is done.

[06:07] Christopher David: It doesn't matter if they put out a llama 4.3. It's like they're leaking now.

[06:15] Christopher David: They're going closed.

[06:22] Christopher David: In many ways, obtaining this goal is a quintessentially American volition. In the face of a technology that is poised to bring such extreme financial and by proxy literal power to a few companies,

[06:32] Christopher David: opening AI is one of the only things we can do to reduce it.

[06:35] Christopher David: Technology proceeds in a one-way direction for a variety of geopolitical and capitalistic reasons. It is impractical to pause AI development to do AI another way.

[06:44] Christopher David: The best we can do is chart a path that makes this future better.

[06:49] Christopher David: Our take is that just AI needs to be decentralized with multiple power centers.

[06:54] Christopher David: There's not going to be zero power centers, but we can at least make sure there's not one or two.

[06:59] Christopher David: We think there should be like a thousand.

[07:05] Christopher David: Do-do-do.

[07:09] Christopher David: We need new systems to mitigate misuse, but it shouldn't be solely up to corporations to control this. Safety by isolating technology to a select few is something we're in the later stages of with nuclear weapons,

[07:19] Christopher David: and AI progress is far harder to monitor.

[07:21] Christopher David: Robustness to AI can only come from designing systems that expect it to be pervasive,

[07:26] Christopher David: not that it is an easy task.

[07:29] Christopher David: Realistically, all of this is fighting gravity.

[07:31] Christopher David: The corporations will win,

[07:32] Christopher David: but we can control to what extent.

[07:34] Christopher David: We can control how good the other options are,

[07:38] Christopher David: the

[07:41] Christopher David: open options.

[07:42] Christopher David: That can be an open corporation too.

[07:45] Christopher David: The call to action here is simple.

[07:47] Christopher David: Consider how you can slightly shift your decision making to make the American deep seek more likely.

[07:53] Christopher David: This approach succeeds just as much by having one model at the end of it as it does by having the community form better habits and norms around the way AI models are conceived,

[08:02] Christopher David: built,

[08:02] Christopher David: shared, and used.

[08:03] Christopher David: Okay, so for my part as a CEO of a tiny American lab,

[08:09] Christopher David: you know, we will support this to the extent that we can.

[08:12] Christopher David: We're more focused on agentic stuff than the base layer,

[08:14] Christopher David: but it sounds like even they acknowledge that the agents are going to be increasingly important.

[08:19] Christopher David: We're basically trying to build the front door of the interface that allows you to very easily use any number of models and agents to pull in different data sets and have it connected to this sort of like decentralized open marketplace that people can get financial contributions for contributing to.

[08:32] Christopher David: uh we're gearing up for like a larger product release later this month so we'll have more on that specifically um you know i'll push back a little bit on the term american i mean i get the point that they're sort of like the you know the different poles of west and east america and china but it more seems like it's the underlying ideal

[08:57] Christopher David: That matters, not the country.

[09:00] Christopher David: And so I hope for this to be a global effort that can engage happily many people other than just Americans.

[09:08] Christopher David: And if we need a better name than American Deep Seek, we can always go with Atlantean Deep Seek.

[09:15] Christopher David: More on that soon.

[09:17] Christopher David: Stay tuned.