Lagrange Point

I wanted to become a theoretical physicist but despite strong ambitions, I couldn't never make it. Yet I keep going back and revisit the fundamental theories describing the world we live in. A thread of my favorite physics books I have read(partly) and you should give it a try 🧵

Machines know everything. The real question is how much do you know ?

Genuinely curious what does Sooryavanshi eat ?

“Stanford, Tesla, OpenAI, Microsoft. And Anthropic now. Reborn in research. Restored to this great gallery of the generative game. A walking work of architecture, vintage, beyond valuation, beyond forgery or imitation. Years since that brilliant researcher first tip-toed onto this storied stage of neural nets. Now, in his immaculate maturity, AK reunited with the frontier.”

Phrases like “The Shashank Redemption” is why I like Alan Wilkins commentary. He has been doing it for more than 30 years now. What an absolute legend.

@arpit_bhayani Curious to understand what are the cross-repo and repo-specific skills you have been using?

The way I test APIs looks so different now :) I make Claude write and run tests. I do not think I, or anyone, will ever write cURL or equivalent commands or use a tool in the near future. My workflow is now simple. Once all the core business logic is checked in and I am ready to test stuff, this is what I do: 1. make claude write tests 2. ensure it covers a good chunk of code 3. make claude write a test spec 4. spec has endpoints, responses, and potential errors 5. claude creates a testing-md file with key details 6. claude uses this file for every change, runs the relevant tests, and this keeps tests updated That is it. For me, claude code handles the mechanics - making HTTP requests, checking status codes, asserting on response shapes. I handle the judgment - whether the test cases actually cover what matters. The split works well because neither side is doing the other's job. (I just realised I am treating it like a co-worker). Given, claude can translate spec or even a thought into test code faster than I can set up any GUI tool, why do things manually :) By the way, the spec also doubles as documentation, which is a free side effect. Whether Claude uses pytest, go test, requests, httpx, or raw cURL under the hood does not matter to me. What matters is that the tests run, fail loudly when they should, and take minutes to write instead of an afternoon. I do not think I will open an API testing tool again. Writing the test logic was always the slow part. That problem is mostly gone now. I am getting AI-pilled, I guess :)

Similarly, an AI model can be trained to think in mysterious and malicious ways. This ML interpretability work is a masterpiece though. Reminds of the developments in nuclear physics in 1930s.

@tsaha You sure, this was not Son of Anton ?

@bryan_johnson Add microbiome in the prompt and let the shrooms kick in

What is the prompt to give Claude a psychedelic experience?

You are always at the center of a technological revolution if you are a part of it.

@aaditsh Your own company

The best companies to work at today: 1. Anthropic 2. OpenAI 3. Google DeepMind 4. SpaceX 5. xAI 6. NVIDIA 7. Tesla 8. Meta 9. Microsoft 10. ?

All roads lead to personalisation.

@NoahKingJr Prompting Claude serendipitously, at will.

TELL ME SOMETHING YOU CAN DO THAT CLAUDE CANNOT

got this beast a week ago

Microbiome should be the word of the year.

@arpit_bhayani Classic race condition :P

arpit1212 is high, I guess :)

I am totally okay with a Goblin on my Mac.

IntelliJ is so insanely well-crafted.

@abhishekcode42 This is also the ReLU activation in neural nets.

@abhishekcode42 Simple. For x<0, in B and D, y =0 and for x >=0, in D, y = x but in B it would be y = 2x. Hence D is the correct answer

Chat, let's solve this in 5 seconds

codex is gobsmaxxed. we are gobsmacked.