Hacker News: makeavish

New comment by makeavish in "GPT-6 Sol and Luna"

makeavish — Wed, 23 Sep 2026 10:38:24 +0000

Don’t think in zero sum terms. OpenAI can’t burn money infinitely, efficient models are better for everyone

New comment by makeavish in "Claude Opus 5.5 Intelligence, Performance and Price Analysis"

makeavish — Tue, 22 Sep 2026 17:31:31 +0000

Nice catch, AA only shows max effort by default and I got disappointed thinking it's a token guzzler though: https://artificialanalysis.ai/models/claude-opus-5-5?models=...

Not sure about how adaptive reasoning works though as they mention adaptive reasoning for every reasoning level

New comment by makeavish in "Claude Status – Elevated errors for multiple models"

makeavish — Tue, 22 Sep 2026 17:23:24 +0000

I follow Leo for the rumors, he has been pretty accurate but again they are rumors so nothing is guaranteed. I don't know his sources. https://x.com/synthwavedd

New comment by makeavish in "Claude Status – Elevated errors for multiple models"

makeavish — Tue, 22 Sep 2026 01:14:46 +0000

There are rumors of Opus 5.5 launching on Tuesday so maybe it due to that? Previously as well there have been incidents just before launch.

New comment by makeavish in "Manus will return to operating as an independent company"

makeavish — Tue, 11 Aug 2026 15:44:05 +0000

No it’s Chinese state who is preventing founders from leaving China https://www.cnbc.com/amp/2026/06/12/meta-reportedly-begins-d...

New comment by makeavish in "HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88"

makeavish — Mon, 29 Jun 2026 18:01:46 +0000

True, I have found it to be valid as well

New comment by makeavish in "HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88"

makeavish — Mon, 29 Jun 2026 11:51:52 +0000

True :(

New comment by makeavish in "HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88"

makeavish — Mon, 29 Jun 2026 09:49:46 +0000

I have been at both side of the market. And it sucks so bad at both ends. Companies which deeply care about next hire are struggling to hire and actual great people looking out are outcompeted by AI slop and AI bulk applying.

It is actually a very hard to solve problem.

New comment by makeavish in "HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88"

makeavish — Mon, 29 Jun 2026 05:41:36 +0000

Hiring and job search has been so hard and AI has amplified the existing problems instead of solving any.

New comment by makeavish in "Claude Agent SDK vs. Messages API for a Production AI Assistant?"

makeavish — Fri, 10 Apr 2026 04:55:39 +0000

Claude launched Managed Agents which fulfills our usecase! https://platform.claude.com/docs/en/managed-agents/overview

New comment by makeavish in "Claude Agent SDK vs. Messages API for a Production AI Assistant?"

makeavish — Wed, 08 Apr 2026 02:43:26 +0000

Interesting, let me check it out

Claude Agent SDK vs. Messages API for a Production AI Assistant?

makeavish — Tue, 07 Apr 2026 02:14:58 +0000

We're building a chat assistant into our B2B developer tool. It needs to call domain-specific tools (MCP), stream responses (SSE), and persist conversation history.

The Claude Agent SDK looks like it would save us a lot of work — agent loop, compaction, hooks, session management all built in. But the more we look at it, the more the architecture feels oriented toward single-user CLI use. The subprocess spawn overhead (~12s per call), filesystem-based sessions, memory issues in long-running processes, and the Node.js subprocess dependency inside our Python backend all makes look like not fit.

The other path is building the agent loop ourselves on the Messages API. Cleaner fit for a web backend, but obviously more work upfront.

Has anyone here gone through this decision? Would especially love to hear from folks who tried the SDK in a multi-tenant setting and can share what worked and what didn't. Also curious about anyone who started with the SDK and migrated away, or vice versa.

Comments URL: https://news.ycombinator.com/item?id=47669957

Points: 2

# Comments: 3

New comment by makeavish in "Gemini 3.1 Pro"

makeavish — Thu, 19 Feb 2026 17:12:16 +0000

Yeah maybe that but atleast let us know about this Or have dynamic limits? Nerfing breaks trust. Though I am not sure if they actually nerf it intentionally. Haven't heard from any credible source. I did experience in my workflow though.

New comment by makeavish in "Gemini 3.1 Pro"

makeavish — Thu, 19 Feb 2026 17:09:59 +0000

I don't use AI studio for my work. I used Antigravity/Gemini CLI and 3 pro was great for few weeks and now it's worse than 3 flash or any smaller model from competitor which are rated lower on benchmarks

New comment by makeavish in "Gemini 3.1 Pro"

makeavish — Thu, 19 Feb 2026 16:44:24 +0000

Looks great!

New comment by makeavish in "Gemini 3.1 Pro"

makeavish — Thu, 19 Feb 2026 16:42:22 +0000

Great model until it gets nerfed. I wish they had a higher paid tier to use non nerfed model.

New comment by makeavish in "Gemini 3.1 Pro"

makeavish — Thu, 19 Feb 2026 16:39:46 +0000

I hope to have great next two weeks before it gets nerfed.

GPT 5.1 research thinks it's 2024 so ignoring search results mentioning 2025

makeavish — Thu, 04 Dec 2025 16:13:44 +0000

Article URL: https://twitter.com/makeavish11/status/1996609547113538039

Comments URL: https://news.ycombinator.com/item?id=46149214

Points: 3

# Comments: 1

ChatGPT prompt consumes equivalent to 10s of Netflix

makeavish — Sat, 29 Nov 2025 06:55:31 +0000

Article URL: https://simonwillison.net/2025/Nov/29/chatgpt-netflix/

Comments URL: https://news.ycombinator.com/item?id=46085641

Points: 8

# Comments: 4

New comment by makeavish in "Google Antigravity"

makeavish — Tue, 18 Nov 2025 20:35:21 +0000

Tried Antigravity for 2 queries and my model quota limit breached. Model definitely felt better than GPT 5.1 (my current daily driver). I am continuing to use Gemini 3 Pro on Cursor to evaluate further.