Indigo Orton
@indigoorton
Followers
47
Following
131
Statuses
182
founder @ https://t.co/b1OaskF0nf | compilers, LLMs, and the future of programming | cambridge phd in program analysis and concurrency
Joined July 2018
Arguably we simply end up in a place where the only software bought is super-high-quality (Eg @linear) because anything less is AI generated internally for a company.
0
0
1
💯 AI-based coding only increases the need for great abstractions. It might not matter for prototyping a web app (though I'd argue the current systems that use @reactjs + @tailwindcss + @supabase are building on great abstractions!), but it is essential for using AI to build sophisticated software. If an AI can code, the next question is "how complex?" and the one after that is "how do you have a human understand/trust what is written" -- that all comes down to managing complexity and confidence. That's abstractions.
0
0
1
HN pattern match: Commenter type 1) I've been using "X" for N months and I love it. I didn't realize Y would be useful, but it's amazing. Type 2) I've heard a lot about "X" and read about it but I really don't get what problem it is solving. Type 2.A) I've read a lot about "X" and it seems interesting but lack of Z is holding me back. Be more experimental and empirical. Learn more.
0
0
0
@0xluffyb Work on a hard problem and identify tech/problems that cracked devs self-select for.
0
0
0
RT @Ar_Douillard: re: OpenAI's DeepResearch on the Last Humanity's Exam on one hand, it's unfair to compare an AI with web-access vs other…
0
19
0
Anyone else getting bugs with Deep Research (@OpenAI's version) not responding after asking for clarification (or after "reasoning")? Had it 3 times in a row now.
0
0
0
@HSVSphere PL you make the calls. So who "cares" what others say (unless it's genuinely persuasive), make the right technical decision. Play PL not Football kids.
0
0
0