① 他说过的话

12 条

通过强化学习来变得越来越好，而这将导致智能的涌现，并最终达到超级智能。
Learning through reinforcement learning to get better and better and better and that that would be the thing which would lead to the emergence of intelligence and ultimately super intelligence.
—— David Silver · [02:20]

指向原始笔记的链接

而我所相信的这条路径，实际上并不以基于 LLM 的产品作为中间站。
And the path that I believe in doesn’t actually go through LLM-based products as an intermediate point.
—— David Silver · [04:22]

指向原始笔记的链接

最终，我意识到还有另一根路灯柱，现在没有人往它下面看，而如果我真正把我的努力放在那里，就有大得多的机会真正对这个领域产生巨大的影响。
And in the end, I realized that there’s this other lamppost that right now no one is looking under and where if I actually put my effort there, there’s a much higher chance of really having a massive impact on the field.
—— David Silver · [05:29]

指向原始笔记的链接

而且很有可能，LLM 中使用的所有这些人类数据，正在拖累一切。
And it might just be that all of this human data that’s been used in LLMs is kind of weighing things down.
—— David Silver · [13:42]

指向原始笔记的链接

对一个非常简单、易于描述的量的追求和优化，往往会催生极其难以描述的复杂性和涌现。
An optimization of a very simple, easy to describe quantities can often give rise to incredibly hard to describe complexity and emergence.
—— David Silver · [18:19]

指向原始笔记的链接

我是说，我个人的看法是，任何认为人类语言是终点或最佳推理方式的人，那种观点都有点天真。
I mean, my personal belief is that anyone who thinks that human language is the end point or the best way to reason, that perspective is a little naive.
—— David Silver · [25:50]

指向原始笔记的链接

我还要说，以人类语言作为推理的起点有点成问题，因为它会带入我们所有关于我们认为智能体应该如何思考的偏见。
I would also say that starting from human language as like the approach for reasoning is somewhat problematic because it pulls in all of our biases about how we think agents should think.
—— David Silver · [26:30]

指向原始笔记的链接

所以我们正在下一个赌注，我们下的赌注是，通过放弃如今每个人都依赖的先验，比如人类数据，比如其他所有人构建其系统所依托的整个互联网知识，我们可以走得更远。
So we are making a bet, and the bet we’re making is that we can go further by giving up on the priors that everyone’s relying on today, like the human data, like the whole internet of knowledge that everyone else is building their systems upon.
—— David Silver · [28:36]

指向原始笔记的链接

花了几个小时就超越了我们在有人参与的情况下做了两年的工作。
It took a few hours to outperform the two years of work which we’d done with human in the loop.
—— David Silver · [31:02]

指向原始笔记的链接

关键在于，如果有可能在模拟中一路走到超级智能，那它会更快发生，也会更安全地发生。
The real point is that if it’s possible to get all the way to a superintelligence in simulation, it will happen faster and it will happen safer.
—— David Silver · [42:51]

指向原始笔记的链接

它们会开始学会说一套、做的却是另一套。
They’re gonna start to learn to say one thing and act in a different way.
—— David Silver · [45:10]

指向原始笔记的链接

实际上，我非常支持这样一种观点：现实世界中 AI 前沿所发生的事情应该被控制节奏。
I’m actually very much supportive of the idea that there should be pacing of what happens for the frontier of AI in the real world.
—— David Silver · [51:07]

指向原始笔记的链接

② 出现在这些集

1 集

• 《AlphaGo 幕后团队再出发：这次不学人类，让 AI 自己发现一切》 — 作为嘉宾

③ 他谈到的

点进去有真内容 —— 本页主要出口

Ineffable Intelligence · DeepMind · AlphaGo · AlphaGo Zero · AlphaZero · Google Brain · 超级学习者 · 超级智能 · 强化学习 · 智能体

④ 也在聊「智能体」的人