François Chollet:从科学发展理解 AI 的递归自我改进
现实世界充满了递归自我改进的系统,但其中一个尤其值得关注:科学,将其视为一个系统(甚至作为一个智能体,具有目标和资源)。如果你想真正理解 AI 的递归自我改进(RSI),科学应该是你的参考点。
科学是一个智能系统,显然它在递归自我改进:
- 科学发现解锁了新技术,帮助构建更好的实验工具。这是几乎所有领域进步的最主要驱动力。
- 它们解锁了新的概念性进步(想法、理论),帮助解决更多问题。
- 它们增加了社会的经济产出,导致更多资源流入科学。
- 它们解锁了更好更快的工具(例如,通过更好的芯片和网络技术获得更多计算能力)。
结果,许多衡量科学输入的指标呈指数增长:
- 人力(每约 15 年翻一番)
- 全球研发支出(翻番速度稍快,每约 13 年)
- 论文和专利(严格来说,这是人力的一种衡量)
- 专用于科学的计算能力(每约 2 年翻一番)
但是,科学进步是指数级的吗?历史上,自工业革命开始以来,科学影响随时间增长的速率大致保持不变(即科学进步是线性的)。1850-1900 年大约和 1900-1950 年或 1950-2000 年一样带来了巨大的变化。
1850–1900:进化论、细菌理论和无菌手术、热力学、电磁场方程、元素周期表、制药、电力、电报和电话、内燃机、摩天大楼、机械化农业……
1900–1950:狭义和广义相对论、量子力学、核裂变和原子能、抗生素、遗传理论、电子计算机、信息理论、合成聚合物和塑料、晶体管、航空……
1950–2000:DNA、遗传工程、集成电路、微处理器和个人计算、互联网、载人航天、登月、卫星通信和 GPS、粒子物理标准模型……
就实际而言,比如预期寿命,自 1840 年以来以每年大约 3 个月的惊人线性方式增加,进步是一条直线。这在影响易于衡量的领域尤其明显,比如生物学、医学和农业。
我最早在 2012 年写过这种现象及其原因,此后几年中,一系列研究证实了这一点。例子包括 Nielsen 和 Collison 的 2018 年论文《科学正变得越来越不划算》,以及 2020 年的经济学论文《想法越来越难找了吗?》(事实上,我相信 Nielsen 的论文源于我六个月前与他就这个确切想法进行的对话)
简而言之,主要原因是研究首先解决最高影响、最简单的问题,而后续每个问题要么更难,要么影响更低。而且是指数级的。提出信息理论的那篇论文写起来并不难(单一作者!),但你很难写出一篇计算机科学论文能在影响力上超过它。
这就是为什么科学作为一个系统需要指数级的资源(输入)来产生线性的影响力(输出)。随着时间推移,它变得指数级地更难。
如果你在思考 AI 的 RSI,这值得深思。我完全相信 AI 的递归自我改进(RSI) 已经在发生,并且未来会加速。但我不相信这会导致“智能爆炸”——那将违背我对智能的一切了解,以及我对递归自我改进系统的一切了解。
对照原文
The real world abounds with recursively self-improving systems, but one in particular deserves attention: Science, modeled as a system (perhaps even as an agent, with goals and resources). If you want to really understand AI RSI, science should be your reference point. Science is an intelligent system, and it is obviously recursively self-improving: 1. Scientific discoveries unlock new technology that helps build better experimental tools. This is a top driver of progress in nearly all fields. 2. They unlock new conceptual advances (ideas, theories) that help solve more problems. 3. They increase society's economic output, leading to more resources flowing into science. 4. They unlock better faster tooling (e.g. more compute via better chip & networking technology). As a result, many measures of scientific *input* grow exponentially: 1. Headcount (doubles every ~15 years) 2. Global R&D spending (doubles a bit faster, every ~13 years) 3. Papers and patents (technically this is a measure of headcount) 4. Compute dedicated to science (doubles every ~2 years) But is scientific progress exponential? Historically, the rate of scientific impact over time has remained roughly constant since the start of the industrial revolution (i.e. scientific progress is *linear*). 1850-1900 was about as dramatic as 1900-1950 or 1950-2000. 1850–1900: Evolution, germ theory & antiseptic surgery, thermodynamics, electromagnetic field equations, the periodic table, pharmaceuticals, electricity, telegraph and telephone, internal combustion engine, skyscrapers, mechanized agriculture... 1900–1950: special and general relativity, quantum mechanics, nuclear fission & atomic energy, antibiotics, genetic theory, electronic computers, information theory, synthetic polymers and plastics, the transistor, aviation... 1950–2000: DNA, genetic engineering, integrated circuits, microprocessors & personal computing, the Internet, crewed spaceflight, moon landing, satellite communications & GPS, standard model of particle physics... In real terms, like life expectancy, which has increased in a remarkably linear fashion of roughly 3 months per year since 1840, progress is a straight line. This is especially apparent for fields where impact is easy to measure, like biology, medicine, and agriculture. I first wrote about this phenomenon and its causes in 2012, and a steady stream of research has confirmed it in the years since. Examples include the 2018 paper by Nielsen and Collison, "Science Is Getting Less Bang for Its Buck," and the 2020 economic paper, "Are Ideas Getting Harder to Find?" (In fact, I believe the Nielsen paper stemmed from a conversation I had with him about this exact idea six months earlier) In short, the primary cause is that research solves the highest-impact, easiest problems first, and every subsequent problem is either harder or lower-impact. Exponentially so. The paper that presented information theory wasn't very hard to write (single author!) but you'd have a hard time ever writing a CS paper that beats it in impact. This is why science as a system requires exponential resources (input) to produce linear impact (output). It gets exponentially harder over time. Worth thinking about if you're pondering RSI for AI. I fully believe AI RSI is already happening and will accelerate in the future. But I do not believe this leads to an "intelligence explosion" -- that would fly in the face of everything I know about intelligence and everything I know about recursively self-improving systems.