duolingo is not a language learning company it is engagement and high retention tech insane DAU powered by streaks and compounding of mechanics you can add anything to duolingo: math, music, drawing - it will work
I’m reading DeepSeek’s new paper and it’s wild. Everyone keeps trying to brute-force reasoning with bigger models and more RL… meanwhile DeepSeek just showed the real bottleneck isn’t model size -- it’s verification. They built a system where the model doesn’t just solve…
the bottleneck of AI = creating verification @_jasonwei is right
All the ambitious people I know end up moving to the US or Dubai. Europe is absolutely cooked.
unpopular opinion: if you're 2 cofounders making $30K a month then it's $15K before fees and before taxes with a 30/40% profit margin it's less than 6K$ per founder before taxes after taxes depending on where you live we can trim another 10-30% of those $6K per founder =…
Any task that's verifiable is easy to train by AI Same progress was in all other fields As soon as verification is easy => models train fast
New blog post about asymmetry of verification and "verifier's law": jasonwei.net/blog/asymmetry… Asymmetry of verification–the idea that some tasks are much easier to verify than to solve–is becoming an important idea as we have RL that finally works generally. Great examples of…
NVDA employees rn 🤣🤣🤣
The richest men in the world are: 75% Entrepreneurs 15% Investors 7% Athletes 3% Artists 0% Employees. Nobody got rich with a salary. Dear son, Be smart enough to realize: 9-5 is a scam.
“be tolerant with others and strict with yourself.” but if you want to win there’s a point you must flip empathetic by default ruthless when the mission is at risk kind heart brutal standards
when NY dudes are getting into AI that means that bubble is coming
AI FOMO in NY is real. A year ago, my NY friends would kinda just laugh at me when I tried to convince them to move to SF, make a joke about homeless people or something. Now, many are planning their moves
great company golden-level material
Growth has been insanely lately... arcads.ai reached $10M arr Here’s the exact playbook we used for each growth stage: $0 → $10k MRR Before writing a single line of code... we went straight to our ICP: We sent simple emails: “Hey, we’re building a tool…
that's actually extremely low
This is wild ONLY 177 apps make more than $5M/mo This is less than 1% of the apps on the App Store To win in the app space you literally have to be part of the 1%
domain knowledge is key actually to guide LLM into the desired direction requires insane domain knowledge as well as doing right evals focus & specialization matters for humans, and it matters for AGI too
The reason nearly all VCs are betting against this is that focus and specialization matters OpenAI launched a direct competitor to Granola months ago and no one cares
why people build businesses with the goal of $5K/mo Go work somewhere for $20K/mo Less risks, better quality of life, you learn faster, you work with the best, you grow faster Indie hacking is for unemployable brokies Build something big or work with someone who build it
United States 트렌드
- 1. Penn State 18.2K posts
- 2. #twitchrecap 12K posts
- 3. Romero 21.6K posts
- 4. #TADCFriend 1,454 posts
- 5. Slay 19.4K posts
- 6. Pat Kraft 2,103 posts
- 7. Paul Dano 1,028 posts
- 8. Zion 9,342 posts
- 9. #GivingTuesday 31.2K posts
- 10. Fulham 44.2K posts
- 11. Lewandowski 27.4K posts
- 12. Larry 60.7K posts
- 13. Adam Thielen 3,827 posts
- 14. Sabrina Carpenter 42.1K posts
- 15. Tarantino 7,322 posts
- 16. Alek Manoah N/A
- 17. Pedri 56.5K posts
- 18. Newcastle 32.6K posts
- 19. Rashford 10.4K posts
- 20. Blanca 30K posts
Something went wrong.
Something went wrong.