Hacker News: Alifatisk

New comment by Alifatisk in "OpenAI Jalapeño: Better than Nvidia Blackwell"

Alifatisk — Tue, 25 Aug 2026 19:39:33 +0000

Why censor yourself?

New comment by Alifatisk in "OpenAI restores 5-hour Codex and Work limits for ChatGPT Plus users"

Alifatisk — Tue, 25 Aug 2026 15:25:28 +0000

Maybe I should go back to ChatGPT again and copy-paste code.

New comment by Alifatisk in "OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)"

Alifatisk — Mon, 24 Aug 2026 16:13:31 +0000

Bummer, this does not affect the weekly usage on Codex through Subscription.

New comment by Alifatisk in "Doomscrolling at work wastes time, but the real cost is what happens after"

Alifatisk — Sun, 23 Aug 2026 10:31:03 +0000

”The effects of consuming negative news during the workday can linger in employees' minds long after they stop scrolling.”

Doomscroll doesn’t necessarily have to be about negative news. I don’t have any social medias but I’ve seen my friends feed, they are pure brainrot.

New comment by Alifatisk in "AI companies destroy physical books – let's scan rare books before it's too late"

Alifatisk — Fri, 21 Aug 2026 06:37:11 +0000

Does it exist in the web at all? Could be the contents have not been scraped for the dataset yet.

New comment by Alifatisk in "Google's AI photoscanner can determine body fat through selfies"

Alifatisk — Thu, 20 Aug 2026 16:28:01 +0000

Can one try out this photoscanner?

New comment by Alifatisk in "Apple announces changes for apps in the European Union"

Alifatisk — Wed, 19 Aug 2026 06:27:06 +0000

Duopoly

New comment by Alifatisk in "Qwen3.8 27B scores 52 on Artificial Analysis"

Alifatisk — Tue, 18 Aug 2026 15:44:00 +0000

Right, I guess the ”Open” in OpenAI means open for speculation.

New comment by Alifatisk in "Qwen3.8 27B scores 52 on Artificial Analysis"

Alifatisk — Tue, 18 Aug 2026 06:36:10 +0000

Is Luna MoE? Had no idea OpenAi went that direction

New comment by Alifatisk in "Research papers using "kidney disappointment" instead of "kidney failure""

Alifatisk — Sun, 16 Aug 2026 13:55:25 +0000

How about "lactose bigotry" instead of "lactose intolerance" https://scholar.google.com/scholar?q=%22lactose+bigotry%22

Research papers using "kidney disappointment" instead of "kidney failure"

Alifatisk — Sun, 16 Aug 2026 12:22:20 +0000

Article URL: https://scholar.google.com/scholar?q=%22kidney+disappointment%22

Comments URL: https://news.ycombinator.com/item?id=49319389

Points: 437

# Comments: 154

Z.ai Security Disclosure

Alifatisk — Fri, 14 Aug 2026 19:22:00 +0000

Article URL: https://cvd.z.ai

Comments URL: https://news.ycombinator.com/item?id=49303433

Points: 32

# Comments: 5

New comment by Alifatisk in "Qwen 3.8 27B"

Alifatisk — Fri, 14 Aug 2026 16:53:16 +0000

Thank you Kenny

New comment by Alifatisk in "Qwen 3.8 27B"

Alifatisk — Fri, 14 Aug 2026 15:40:01 +0000

> I'd gladly take A5B or A8B or even A10B as a sort of middle ground.

Whats up with focusing on the active param count? Do yall fiddle with the weights or something?

New comment by Alifatisk in "Show HN: MCP Memory – Fast Agent Memory Using Google's OKF and SQLite FTS5"

Alifatisk — Fri, 14 Aug 2026 09:25:59 +0000

Didn’t know I was basically replicating AGENT.md. I didn’t jump on that train when it became a trend.

New comment by Alifatisk in "GLM-5.3: Frontier coding with emergent cyber capabilities"

Alifatisk — Fri, 14 Aug 2026 06:31:10 +0000

We’ll see, the best benchmark is your own. Looking forward to try this out!

New comment by Alifatisk in "Gemini 3.7 Flash"

Alifatisk — Thu, 13 Aug 2026 19:11:02 +0000

I am not sure this enlightens you with anything but I have a TODO.md file with three headlines. Todo, Doing and Done. The agent is aware of it and knows on which task we are on.

On complete, it moves the user story from Doing to Done. I also have a MEMORY.md file that the agent read and writes in the beginning of a new conversation and at the end of our conversation to update stale information. These files are referred to every time I start a new conversation.

Regarding forking, I know Codex has such button underneath each message that lets you fork the whole conversation. I usually do that when I want to sidetrack and discuss something.

I use no SKILLS or commands like /goal. I’ve come a long way with just prompts and markdown files. Its all a different way of encapsulating instructions anyways.

New comment by Alifatisk in "Gemini 3.7 Flash"

Alifatisk — Thu, 13 Aug 2026 19:01:02 +0000

Does it reset at every turn? From my experience in Codex for example, Luna (Max) fills the 256k token window relatively quick. The only thing lowering the context window again is the compaction.

New comment by Alifatisk in "Show HN: MCP Memory – Fast Agent Memory Using Google's OKF and SQLite FTS5"

Alifatisk — Thu, 13 Aug 2026 18:40:38 +0000

What I do currently is having a markdown file named MEMORY.md at the root folder. Then, whenever I create a new conversation with an agent, I refer to that file. That file becomes the initial source of context and knowledge. At the end of an task, before I leave the conversation, I ask the agent to update MEMORY.md with lessons and new knowledge it has gathered from our conversation, it also removes stale or outdated information from that markdown file.

I myself do not care what's written in that file, I steer, instruct and share my knowledge, visions, goal and preferences in our conversations, and the agent will boil that down and update the markdown folder. It has worked very well for me.

I now do not have to worry about creating handoff prompts when creating a new conversation or that I have to teach an agent from the ground up about the context we're in, I just refer to that markdown file.

New comment by Alifatisk in "Gemini 3.7 Flash"

Alifatisk — Thu, 13 Aug 2026 18:09:07 +0000

Ever since the insane discount with GPT-5.6 Luna, not much excites me anymore. I mean just look at the benchmarks, even though Gemini 3.7 Flash performs well on the DeepSWE 1.1, Luna (Max) still performs way better. I personally have stuck to Luna (Xhigh) because its been more than enough and does not bloat up the context window too fast with reasoning tokens.

https://deepswe.datacurve.ai

> Starting January 1, 2027, $1.50/1M input tokens and $7.50/1M output tokens will apply.

Compare this to Luna which is at $0.2/1M input ($0.02 cached) and $1.2/1M output.

https://developers.openai.com/api/docs/models/gpt-5.6-luna