Hacker News: bobkb

New comment by bobkb in "Jamesob's guide to running SOTA LLMs locally"

bobkb — Fri, 03 Jul 2026 18:23:02 +0000

Very useful. The whisper setup is something similar to what we have been using. The LLM setup though is outstanding.

New comment by bobkb in "Markets are competitive if and only if P != NP"

bobkb — Fri, 03 Jul 2026 16:29:30 +0000

This gave me shivers

New comment by bobkb in "Previewing GPT‑5.6 Sol: a next-generation model"

bobkb — Sat, 27 Jun 2026 06:25:04 +0000

Will it be accessible to anyone ?

New comment by bobkb in "Anatomy of a Failed (Nation-State?) Attack"

bobkb — Sat, 27 Jun 2026 06:20:59 +0000

This type of attack is going on for few years now. I had 2 in my credit.

Some details https://freebird.in/malicious-code-source-code-shared-via-jo...

New comment by bobkb in "SpaceX to buy Cursor for $60B"

bobkb — Tue, 16 Jun 2026 17:00:45 +0000

Will cursor launch a CLI tool like Claude/codex/opencode/pi ?

New comment by bobkb in "Formal methods and the future of programming"

bobkb — Mon, 15 Jun 2026 15:34:00 +0000

I assume your idea is, if the spec and the proof is verified the code generated is good enough as well ?

New comment by bobkb in "Formal Methods and the Future of Programming"

bobkb — Sun, 14 Jun 2026 15:08:52 +0000

I have been testing formal verification methods with multiple products. It will be great to also understand more about what’s tried and how it was done. For example attempting to verify the spec is what I have been trying to implement.

New comment by bobkb in "Claude Fable 5"

bobkb — Tue, 09 Jun 2026 18:25:13 +0000

In an interesting coincidence I ended up watching Person of Interest S4 E5 while reading the announcement. The series showed some code supposedly belonging to to an AI.

Fable 5 said the first screen shot is from “ IDA Pro’s Hex-Rays decompiler” and a windows driver. The second screenshot triggered the safety guard rails and pushed me into Haiku.

Apparently the code is Windows driver code.

New comment by bobkb in "LLMs are eroding my software engineering career and I don't know what to do"

bobkb — Sun, 07 Jun 2026 16:18:34 +0000

It’s impossible to write a spec that’s not ambiguous , complete and correct in natural languages. Thus prompts will always generate unreliable software.

New comment by bobkb in "LLMs are eroding my software engineering career and I don't know what to do"

bobkb — Sun, 07 Jun 2026 16:11:19 +0000

IMHO even if we are using auditing tools I believe we must use deterministic tools for critical analysis like this. Such rule and pattern based systems may not scale beyond certain point but they can be accurate.

New comment by bobkb in "I design with Claude more than Figma now"

bobkb — Sun, 07 Jun 2026 10:20:46 +0000

At work we are now in the process of migrating away from Figma. We had spend years perfecting our Figma based design workflow. Currently we are moving all the designs into the code itself using Storybook. The gap currently is reviews and feedback which is addressed by Chromatic now.

New comment by bobkb in "Ask HN: What was your "oh shit" moment with GenAI?"

bobkb — Sat, 06 Jun 2026 14:26:13 +0000

I tried building a deliberately vague project around managing MCP servers [0]. The purpose was to find what LLMs and agents can do. While the project didn’t reach anywhere I was amazed by how it’s possible to navigate even with no clear direction. The ability of the “glorified auto-complete” system to pull off something this sort was an eye opener for me.

0. https://github.com/bobinson/aop1

New comment by bobkb in "Open Code Review – An AI-powered code review CLI tool"

bobkb — Fri, 05 Jun 2026 11:02:51 +0000

False positives from the deterministic audits a very difficult problem to address. Comparing and deduplicating across different methods or LLM audits seems to the only way.

New comment by bobkb in "Anthropic's open-source framework for AI-powered vulnerability discovery"

bobkb — Thu, 04 Jun 2026 22:08:09 +0000

I think these audit tools can look beyond just security and can look for compliance audits as well. The ability to audit real targets in staging environments makes it easy to identify issues.

New comment by bobkb in "Anthropic's open-source framework for AI-powered vulnerability discovery"

bobkb — Thu, 04 Jun 2026 22:05:56 +0000

Very interesting.

I have working on and using a similar tool for a while now :

https://github.com/bobinson/vulture

I have been struggling with false positives and using Claude + MCP as a poor man’s audit tool. As of last few days found better result with nvidia hosted models.

New comment by bobkb in "Malicious npm packages detected across Red Hat Cloud Services"

bobkb — Mon, 01 Jun 2026 14:08:10 +0000

When will npm issues stop ? This has become a big pain !

New comment by bobkb in "DeepSeek reasonix, DeepSeek native coding agent with high caching and low cost"

bobkb — Sun, 24 May 2026 19:17:03 +0000

That’s impressive!

On the sheer performance it’s comparable to Opus ?

New comment by bobkb in "SANA-WM, a 2.6B open-source world model for 1-minute 720p video"

bobkb — Sat, 16 May 2026 17:14:21 +0000

Agreed the marketing angle. But beyond the marketing angle what seems to matter is the access to data - look at Seedance , various Kling models etc which are far ahead of others.

New comment by bobkb in "SANA-WM, a 2.6B open-source world model for 1-minute 720p video"

bobkb — Sat, 16 May 2026 13:01:36 +0000

The trouble is the lack of training available to these models compared to the ones like Seedance and Kling who seems to be tapping into their unlimited video inventory. Many models like LTX is technically good but when it comes to slightly different camera movements or the subject interacting with objects they struggle. For a recent example we had to use sample videos generated by closed source models and then use the same for final video.

New comment by bobkb in "The Boring Part of Bell Labs (2025)"

bobkb — Wed, 13 May 2026 12:19:51 +0000

Nostalgic alumni here :)