Hacker News: laurentiurad

New comment by laurentiurad in "We're testing new ad formats in Search and expanding our Direct Offers pilot"

laurentiurad — Thu, 21 May 2026 11:13:50 +0000

What do you expect them to say? More annoying ads? They're trying to wrap this in a positive way. Everyone knows that ads are annoying.

New comment by laurentiurad in "Agents can now create Cloudflare accounts, buy domains, and deploy"

laurentiurad — Wed, 06 May 2026 09:23:58 +0000

Excellent tool for scammers.

New comment by laurentiurad in "OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors"

laurentiurad — Tue, 05 May 2026 07:44:50 +0000

I cannot agree more. Useful idiots and people working in private equity which have a direct financial interest to hype this tech.

New comment by laurentiurad in "OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors"

laurentiurad — Tue, 05 May 2026 07:10:13 +0000

It's tech bros like you that are to blame for the shortage of radiologists supporting statements since 2016 which state that radiologists will disappear. Your great SOTA LLMs will tell you to walk 5 mins to the car wash instead of taking the car.

New comment by laurentiurad in "OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors"

laurentiurad — Mon, 04 May 2026 07:55:04 +0000

Good luck to you if the prompt is written by health insurance.

New comment by laurentiurad in "OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors"

laurentiurad — Mon, 04 May 2026 07:54:15 +0000

next year bro, I promise, now give me 60 billion more in funding

New comment by laurentiurad in "OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors"

laurentiurad — Mon, 04 May 2026 07:51:44 +0000

You are confusing the job with a subset of tasks. Some tasks can be automated, some won't. That doesn't mean LLMs, which cannot tell how many r's are in strawberry, will replace anyone.

New comment by laurentiurad in "Amateur armed with ChatGPT solves an Erdős problem"

laurentiurad — Sun, 26 Apr 2026 12:28:07 +0000

This program was brought to you by the private equity engagement pod.

New comment by laurentiurad in "DeepSeek v4"

laurentiurad — Fri, 24 Apr 2026 09:08:47 +0000

Try Opencode or Comrade. Both OSS and working great with OSS models too.

New comment by laurentiurad in "DeepSeek v4"

laurentiurad — Fri, 24 Apr 2026 08:59:57 +0000

not a full AI stack. Training still runs on NVIDIA chips.

New comment by laurentiurad in "Show HN: I built Comrade – the security-focused AI agent"

laurentiurad — Mon, 20 Apr 2026 20:04:11 +0000

I designed Comrade to have a "common sense" engine, where each plan/action goes through a filter before being suggested to the user to be approved. This is particularly enforced when the agent goes on a web page. It will always be aware of the source of the prompt, and if it's different from the Electron app where the user can interact with the agent, it will drop that instruction altogether.

Show HN: I built Comrade – the security-focused AI agent

laurentiurad — Mon, 20 Apr 2026 19:42:00 +0000

Inspired by the success of OpenClaw, I built Comrade, which is an open-source AI workspace for teams focused on security. It provides a premium interface for AI-powered workflows, built with transparency, extensibility, and local-first principles. Check it out here: https://github.com/LaurentiuGabriel/comrade

Comments URL: https://news.ycombinator.com/item?id=47839520

Points: 5

# Comments: 2

New comment by laurentiurad in "Show HN: I Recreated Encarta's MindMaze"

laurentiurad — Mon, 20 Apr 2026 13:30:47 +0000

Thanks for the feedback!

New comment by laurentiurad in "Show HN: I Recreated Encarta's MindMaze"

laurentiurad — Mon, 20 Apr 2026 10:08:35 +0000

Thanks a lot!

I have created my own engine for testing the simulations in TypeScript. There is no external logic circuit library under the hood.

The simulation itself is tick-based with a light event-driven flavor. Each tick I compute an evaluation order using a topological sort of the circuit graph (based on wire connections). Then I iterate through components in that order, gathering inputs from connected wires. Each component runs an evaluate() function, producing outputs and a nextState. Outputs propagate through wires immediately within the same tick.

New comment by laurentiurad in "Show HN: I Recreated Encarta's MindMaze"

laurentiurad — Mon, 20 Apr 2026 10:01:56 +0000

Thank you so much! I have this on my list, so I will try to add this tonight.

New comment by laurentiurad in "Show HN: I Recreated Encarta's MindMaze"

laurentiurad — Mon, 20 Apr 2026 09:16:52 +0000

In the article I have also talked about a different game as well that teaches players how chips work. I have added that here too: https://select.supply/game/chipbuilder.

Show HN: I Recreated Encarta's MindMaze

laurentiurad — Mon, 20 Apr 2026 08:59:25 +0000

You can check the game here: https://select.supply/game/mindmaze. Curious to learn your feedback. I personally loved the game so much!

Comments URL: https://news.ycombinator.com/item?id=47831807

Points: 6

# Comments: 8

New comment by laurentiurad in "I revived Encarta's Mindmaze and added a new game teaching how to build chips"

laurentiurad — Mon, 20 Apr 2026 08:07:02 +0000

I revived Encarta's Mindmaze and added a new game teaching how to build chips on my affiliate website select.supply. The games are 100% free and no account is needed. I used to spent lots of hours on Mindmaze and just wanted to relive that experience again.

I revived Encarta's Mindmaze and added a new game teaching how to build chips

laurentiurad — Mon, 20 Apr 2026 08:07:02 +0000

Article URL: https://laurentiu-raducu.medium.com/i-added-games-to-select-supply-and-heres-why-bc6db06f8bd4

Comments URL: https://news.ycombinator.com/item?id=47831527

Points: 2

# Comments: 1

New comment by laurentiurad in "The Closing of the Frontier"

laurentiurad — Sun, 12 Apr 2026 19:57:27 +0000

My bet that this is just a model with a slimmer safety alignment which yields results in 100 agent running in parallel configuration.