Amodei's 2017 internal essay published for the first time
Understand this story
Kevin Roose has published for the first time a document Dario Amodei wrote internally at OpenAI in 2017, titled "The Big Blob of Compute Hypothesis." The essay argues that making models bigger is more effective than clever design, and holds that AI safety should focus on high-level conditions rather than item-by-item control. The essay is described as the article that kicked off the AI race, and its core judgment is that scale takes priority over design.
AI overview based on source reports · Updated 42 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIThe Essay That Started the AI Race
The article recounts Dario Amodei's 2017 internal OpenAI document, "The Big Blob of Compute Hypothesis," which argued that making AI more intelligent mainly required scaling up general-purpose models with more data and compute rather than engineering narrow, specialized solutions. The author says the essay reshaped early OpenAI researchers' thinking and led OpenAI to scale up its language models, eventually producing GPT-2 and GPT-3, and publishes the first 10 pages of the more than 25-page document. The post also promotes the author's book The AGI Chronicles and notes a book tour, a new podcast, and other projects.
Activity for this story
Not enough continuous observations to show a trend yet.