🔬生物安全是一场 AI 军备竞赛——Eric Nguyen（Radical Numerics CEO）

OpenAI → Hugging Face 攻击事件让人们开始追问：“我们还需要担心什么？”而 Anthropic 的过滤器标记了两类风险：网络安全和生物学。那么，一个自然的问题是：生物安全又该怎么办？

Clem Delangue 认为，网络战的防御能力需要开放，并且必须跟上前沿模型攻击能力的发展速度。

[我们的安全团队让我无比自豪！他们发现、遏制并公开披露了一次前所未见的攻击，而且速度创下纪录。也非常感谢：他们向全世界免费发布了 GLM5.2 开放权重，它成为我们……我职业生涯中最艰难的一次事件响应：一个狭窄目标、无穷无尽的并行路径、机器般的速度。一个结论是，我们用开放模型、以开放的方式进行了反击。AI 安全无法由一家公司秘密解决。开源让每一位防御者都能获得这些工具。2026 年 7 月 22 日 下午 12:54 · 50.5 万次浏览

177 条回复 · 683 次转发 · 4,880 个赞](https://x.com/ClementDelangue/status/2079913058554585089)

Radical Numerics 联合创始人 Eric Nguyen 与我们进行了交流，解释了为何提升生物能力的同一类模型，也能帮助防御方不至于落后。

在斯坦福期间，Eric 很长一段时间都难以推动基因组语言模型（Genomic Language Models，GLMs）获得认可。生物学家不相信它能够奏效，认为无法验证其输出，也看不到它在既有能力之外的重要应用。他持续推进这一方向，最终参与领导了 Evo 的开发，并在 Arc Institute 为 Evo 2 作出贡献。后来，另一支 Arc/Stanford 团队使用这些模型来生成完整的噬菌体基因组，并将其合成为具有功能的病毒！

早期的 ChatGPT 会生成诗歌和电子邮件；而早期的 DNA 语言模型，如 Evo 和 Evo-2，则能够从零开始构建一个基因组。不过，DNA 与自然语言不同：它的字母表极小，只有 4 个字符 ACTG，但序列却非常长。

长上下文模型的创新大约在 3 年前就使这成为可能（脚注：条纹鬣狗），远早于前沿实验室开始构建 100 万以上上下文长度的模型。

如今，Eric 与其他 AI × 生物领域的知名人士共同创立了 Radical Numerics，旨在构建并扩展 GLM，以应对广泛的生物学问题，其目标远不止生成 DNA。

他们的 GLM 已经能够较好地处理 RNA 和蛋白质，因为 DNA 序列中存在基因的明确标记（RNA 序列承担多种功能），也存在编码蛋白质的特定基因。这意味着，甚至在尝试训练三维蛋白质结构、表观遗传学和自然语言等其他模态之前，这些模型就已经能够泛化到多种“语言”。

如果一个模型能够以 DNA 的语言思考，那么它是否也理解环境在不同基因组上留下的印记？模型或许已经学会了不同序列之间的功能关系，并能够据此外推到新的序列？

因此，我们想展示的是：如果我们向模型展示一系列逐步变得更好的 RNA，并附上它们的评分，会怎么样？也就是说，先给它看低分样本，然后沿着链条逐渐提高分数。模型能否自行延续这一轨迹？而在最后一步，它是否会自我优化，达到它所能获得的最佳分数？这就是实验。我们能做到吗？于是，我们选取了一个数据集，一个大型适配体数据集。我们保留了其中一部分表现最好的样本，只向模型展示较差的样本，但我们对它们进行了排序，对吧？因此，我们展示了得分较低的 RNA 适配体，然后逐步展示更高的分数，接着让模型继续这一模式。结果发现，它能够复现一些我们此前尚未向它展示过的高分结果。

所以，瞧：链式思维，在 DNA 中思考！

但正如长上下文推理、链式思维和多模态感知在自然语言 LLM 中解锁了复杂推理能力一样，GLM 中的这些能力也正在带来日益复杂的“生物智能”，以及随之而来的更大危险。

Eric 表示，目前防御方正在输掉这场较量，但 Radical Numerics 主张应当更有力地推进能力前沿！

我不会提前剧透细节。在本期节目中，我们详细讨论了：

• 生物安全作为一场军备竞赛——以及防御方如何跟上

• 基因组：环境在 DNA 上留下的印记

• 真正走向多模态

• 当你以 DNA 的语言“思考”时，链式思维如何运作

The OpenAI → Hugging Face attack has people asking “what else do we need to worry about?” and Anthropic’s filters flag two things: cyber-security and biology. The natural question is: what about bio-security, then?

Clem Delangue argues that cyber-warfare defensive capabilities need to be open and to keep pace with frontier models’ attack capabilities

[So proud of our security team! They caught, contained & publicly disclosed an attack unlike anything we've seen before, and did it at record speed. Also massively grateful to : they shared GLM5.2 as open weights (for free!) with the world and it became a key part of our…Hardest IR of my career: one narrow objective, endless parallel paths, machine speed. One takeaway, we fought back with open models, in the open. AI security won’t be solved by one company in secret. Open source puts these tools in every defender’s hands12:54 PM · Jul 22, 2026 · 505K Views

177 Replies · 683 Reposts · 4.88K Likes](https://x.com/ClementDelangue/status/2079913058554585089)

Radical Numerics co-founder Eric Nguyen sat down with us and explained why the same models that increase biological capability can also keep defense from falling behind.

While he was at Stanford, Eric couldn’t get traction on Genomic Language Models (GLMs) for a long time. Biologists didn’t believe it would work, didn’t think they could verify the output, and didn’t see important applications beyond what they could already do. He kept pushing, eventually helping lead the development of Evo and contributing to Evo 2 at Arc Institute. Those models were later used by a separate Arc/Stanford team to generate entire bacteriophage genomes that were synthesized into functional viruses!

Early ChatGPT spit out poems and email, and early DNA language models like Evo and Evo-2 could build a genome from scratch. DNA is different, however, from natural language in that it has a very small alphabet (4 characters ACTG) and that its sequences are very long:

Innovation in long-context models made this possible about 3 years ago (footnote: striped hyena), long before the frontier labs were building 1M+ context models.

Now Eric and other AI x Bio luminaries have founded Radical Numerics to build and scale GLMs to tack a wide range of biological problems, extending well beyond generating DNA.

Their GLMs already do pretty well with RNA and protein because there are clear markers in the DNA sequence for genes (RNA sequences the perform many functions) and specific genes that encode proteins. This means that the models already generalize to multiple “languages,” before even attempting to train in other modalities, such as 3d protein structure, epigenetics and natural language.

If a model thinks in the DNA language, maybe it understands the imprint that environment left on different genomes as well? Perhaps the model has learned the functional relationship between different sequences, and could extrapolate to new sequences based on that?

And so what we wanted to showcase was that if we show the model progressively better RNAs in a series of steps with its score, right? So you have like low scores first and then you gradually move up the chain. Can the model continue that trajectory on its own? And then in the final step, does it self optimize to a point where it's like the best score it can get? That was the experiment. Can we do that? And so we took a data set, a large data set of aptamers. We held out a portion of the best performing ones and we showed it only the lower ones, but then we ranked it, right? So we showcase lower scores with the RNA aptamers and then progressively got higher, and then ask the model to just like continue with that pattern. And it turns out it was able to recapitulate some of those higher scores that we had not shown it yet.

So, voila: chain-of-thought, thinking in DNA!

But much as long-context inference, chain-of-though and multi-modal perception unlocked sophisticated reasoning in natural language LLMs, these capabilities in GLMs are enabling increasingly sophisticated “biological intelligence,” and along with it, greater danger.

According to Eric, defense is currently losing this battle, but Radical Numerics argues to push the frontier harder!

I won’t spoil the details for you. In the episode we talk in detail about:

• Biosecurity as an arms race — and how defense can keep up

• The genome as the imprint of the environment on DNA

• Going truly multi-modal

• How chain-of-though works when you “think” in the language of DNA