HNHacker News
TopNewBestAskShowJobs

echen

1,299 karma · joined August 1, 2012

Edwin Chen - founder at Surge AI (https://www.surgehq.ai), next-gen data labeling platform for next-gen AI

Ex Google, Facebook, Twitter, Dropbox, Microsoft Research, MIT.

edwin[æ]surgehq.ai

https://www.surgehq.ai/blog

https://blog.echen.me

https://twitter.com/echen

https://github.com/echen

submissionscomments

Inverse scaling prize for finding a task where larger language models do worse

github.com·3 pts·echen·
0

‘Anti-hunger’ molecule forms after exercise, scientists discover

med.stanford.edu·289 pts·echen·
268

Book Review: San Fransicko

astralcodexten.substack.com·4 pts·echen·
0

The delivery apps reshaping life in India’s megacities

technologyreview.com·2 pts·echen·
0

Meta Made Millions from Ads That Spread Disinformation

wired.com·7 pts·echen·
1

Humans vs. Gary Marcus: The Complexity of Measuring Machine Intelligence

surgehq.ai·3 pts·echen·
0

Meta has built a new language AI–and it’s giving it away for free

technologyreview.com·3 pts·echen·
2

Senate bill would ban data brokers from selling location and health data

arstechnica.com·3 pts·echen·
0

On the Advance of Making Language Models Better Reasoners

arxiv.org·2 pts·echen·
0

How We Built It: OpenAI's GSM8K Dataset of 8,500 Math Problems

surgehq.ai·4 pts·echen·
0

A founder’s guide for picking the right startup law firm

carta.com·2 pts·echen·
0

A Face Search Engine Anyone Can Use Is Alarmingly Accurate

nytimes.com·7 pts·echen·
6

Sentiment Analysis Dataset of Social Media Stock Conversations

surgehq.ai·2 pts·echen·
0

Are the Spammers Winning? Failures in Gmail Spam Detection

surgehq.ai·10 pts·echen·
0

Restoring and attributing ancient texts using deep neural networks

nature.com·1 pts·echen·
0

We asked 100 humans to draw the DALL·E prompts

surgehq.ai·138 pts·echen·
73

Use NLP to tackle online toxicity now

txt.cohere.ai·3 pts·echen·
0

Chronicles of Opt-175B Training

github.com·1 pts·echen·
0

Anthropic Raises Series B to Build Steerable, Interpretable, Robust AI Systems

anthropic.com·3 pts·echen·
1

Nat (Unit)

en.wikipedia.org·1 pts·echen·
0

Risks from Learned Optimization in Advanced Machine Learning Systems

arxiv.org·2 pts·echen·
0

Lessons Learned on Language Model Safety and Misuse

openai.com·11 pts·echen·
1

Fine-Tuning OpenAI's GPT-3 (Weights and Biases)

youtube.com·2 pts·echen·
0

What if social media optimized for human values? A Facebook case study

surgehq.ai·12 pts·echen·
1

Are popular toxicity models simply profanity detectors?

surgehq.ai·183 pts·echen·
211

Is Google Search Deteriorating? Measuring Google's Search Quality in 2022

surgehq.ai·470 pts·echen·
414
← PreviousPage 2 of 2