@rohanpaul_ai: Google's "Attention is All You Need" paper came from trying to get a 3% gain in Google Translate. Innovation is a conse…
Summary
A tweet highlights that Google's seminal 'Attention is All You Need' paper originated from a modest attempt to improve Google Translate by 3%, illustrating that innovation often arises from production challenges.
View Cached Full Text
Cached at: 05/16/26, 05:22 PM
Google’s “Attention is All You Need” paper came from trying to get a 3% gain in Google Translate.
Innovation is a consequence of production. “If you don’t make the thing, you cede your opportunity to innovate on the thing.”
~ Palantir’s CTO @ssankar https://t.co/pltzfau4Pg
Similar Articles
@reach_vb: Attention truly is all you need
A playful tweet referencing the famous "Attention Is All You Need" transformer paper.
@FuSheng_0306: In an interview with Yao Shunyu, Google's internal strategy is indeed going all out to catch up. Google had been competing with OpenAI on chatbots, and fortunately, Gemini 3 performed well, increasing its market share. However, the rise of Anthropic made Sergey Brin realize that the decisive battle of large models lies in code-writing ability...
The article discusses Google's internal strategic adjustment in the face of competition from OpenAI and Anthropic. Google saw some success with Gemini 3, but realized the decisive battle of large models is in code-writing ability, reflecting the urgency of catching up.
@dosco: i'm seeing a lot of industry papers that are karpathy's auto research loop (not cited) or a codex optimization goal for…
A critical observation about recent industry AI papers lacking novelty, citing examples like SkillOpt that treat natural-language skills as trainable external parameters.
@zarazhangrui: If you've adopted AI at your company but haven't seen any tangible results, read this 1990 article: "The Dynamo and the…
This tweet draws a parallel between the slow productivity gains from early electricity adoption and current AI adoption, arguing that true benefits come from redesigning workflows rather than simply bolting AI onto existing processes. It references Paul David's 1990 article 'The Dynamo and the Computer'.
@Raytar: he tested 5760 architectures at Google for a full year. the winner was the original Transformer from 2017. Hyung Won Ch…
Hyung Won Chung shares at MIT that after testing 5760 architectures at Google, the original 2017 Transformer was the best, then he moved to OpenAI to train o1. He claims 99% of AI research is theater.