HNHacker News
TopNewBestAskShowJobs

hemkeshr

3 karma · joined June 17, 2026

submissionscomments
hemkeshr··on Agent memory needs its OpenTelemetry moment
I agree this is very much needed but I doubt the other ai companies would a-comply with this as they need users and mainly paid users otherwise they'll lose themselves. What I think is that with 20$ subs of 2-3 ai can give you more usage compared to their 100$ subs if you do it that way. I don't know will these companies work together to form a standardized memory system. I highly doubt that.
hemkeshr··on Gemini Omni 1.2 Flash
yep they need to improve their harness and their ai ecosystem and make it simpler. They have great ai system but the data retention policy is very bad if you turn it off you lose all of your previous chat data. They need to think from the point of customers nit only profits.
hemkeshr··on Gemini 4 Carbon
It’s still not ready for public—looks like they’re pushing the launch, kind of like they did with argon. Now anthropic and openai will launch a model outperforming carbon then google thinks of calling carbon to public. Also their premium plans are very bad, least usage for the money invested.
hemkeshr··on OpenAI restores 5-hour Codex and Work limits for ChatGPT Plus users
Imagine being a student and using codex for some night studies and hit the limit at the wrong time saying let's continue after exams, still we have normal chat available. Maybe for projects also but we still have no limits in normal chat section but that ain't codex for working on our projects.
hemkeshr··on DeepSeek V4 Pro 0813
I had high expectations for V4 Pro, especially since DeepSeek V4 Flash 0731 performed so well compared with other Flash models. What a letdown.
hemkeshr··on GLM 5.2 Performance Benchmarks
Local models are already useful today. The next milestone is getting this level of performance onto truly affordable hardware.