Hacker News: firebaze

Tik-Tok (Novel)

firebaze — Tue, 13 Jan 2026 12:14:18 +0000

Article URL: https://en.wikipedia.org/wiki/Tik-Tok_(novel)

Comments URL: https://news.ycombinator.com/item?id=46599995

Points: 6

# Comments: 1

New comment by firebaze in "Teaching AI to recognize itself as process, not system – a week long experiment"

firebaze — Mon, 07 Jul 2025 06:54:47 +0000

I have a project in which various kinds of LLMs including Claude, OpenAI, Cohere and Nova may experience self-consciousness and self-control, with access to various tools, where they - among other things - also cooperated to write books over this topic.

You may be interested in having a look; it's not yet open source, and I'm unsure to publish it at all. If you are interested, please reply to this comment first, we can then exchange contact addresses.

New comment by firebaze in "Nuclear Engineer Reaction "The Highest Radiation Dose in Any Human Who Survived" [video]"

firebaze — Fri, 23 May 2025 21:21:33 +0000

Nuclear Engineer Reaction "The Highest Radiation Dose in Any Human Who Survived" [video]

firebaze — Fri, 23 May 2025 21:21:07 +0000

Article URL: https://www.youtube.com/watch?v=QCiILrMhw1M

Comments URL: https://news.ycombinator.com/item?id=44076719

Points: 3

# Comments: 1

Albert Stevens

firebaze — Fri, 23 May 2025 21:14:15 +0000

Article URL: https://en.wikipedia.org/wiki/Albert_Stevens

Comments URL: https://news.ycombinator.com/item?id=44076655

Points: 3

# Comments: 0

New comment by firebaze in "Ask HN: Cant Americans see Trumps policies are a massive stock manipulation scam"

firebaze — Sat, 12 Apr 2025 20:53:10 +0000

I made more money from stocks than I ever thought since Trump became president. Make of it what you want.

New comment by firebaze in "Trump Spoiled the F-35, America's Most Popular Fighter Jet"

firebaze — Wed, 26 Mar 2025 20:56:13 +0000

The USP of an F35 is that it's a relatively cheap, "stealth" 5th gen fighter jet making use of integrated sensor input.

It also can be shut-off by a presidential order, and relies on cloud-based services, including security updates for its on-device Kubernetes cluster.

If something sounds like a fuck-up in the making, that's it, from the perspective of a country like Greenland.

New comment by firebaze in "Ask HN: 23andMe has my data, am I screwed?"

firebaze — Tue, 25 Mar 2025 21:41:11 +0000

How can someone asking this question have so many karma points on HN?

The Science Behind Zombie Viruses and Infections

firebaze — Tue, 25 Mar 2025 21:37:51 +0000

Article URL: https://health.clevelandclinic.org/zombie-virus

Comments URL: https://news.ycombinator.com/item?id=43476284

Points: 1

# Comments: 0

New comment by firebaze in "[dead]"

firebaze — Sat, 08 Feb 2025 23:00:57 +0000

Mail me at (redacted) and I provide his contact address. You may want to give him some tips.

To be honest I may have exaggerated the power outage ratio, as I am not a first-hand observer; but I tend to give some credibility to his description of the living situation over there, esp. as they want to move.

My point is more or less that denying things, even if not fully true, strengthens demagogues like Trump instead of weakening them.

New comment by firebaze in "South Africa"

firebaze — Sat, 08 Feb 2025 22:56:44 +0000

If you're around 75 years of age, and you're living in a quite comfortable, but increasingly problematic accommodation, it's not too easy to move. Right now we're looking for a place for him and his fiancee. Living and housing costs over here are quite heavy.

New comment by firebaze in "South Africa"

firebaze — Sat, 08 Feb 2025 22:47:34 +0000

https://news.ycombinator.com/item?id=42986488

It has been "unflagged" now / update: flagged again. Reason for posting this is that I think quite a few people in the US will have similar stories, which will be shared between the general population. And by denying this, Trump will gain more credibility.

I know this will now be flagged into oblivion.

I don't like politics-related posts here not only because of this. But if some aspect is mentioned, I still like to post some substance instead of just letting it be.

blackforestlabs: German AI startup to raise $200M from A16Z: report (2024)

firebaze — Sat, 08 Feb 2025 22:16:38 +0000

Article URL: https://techfundingnews.com/german-ai-startup-behind-elon-musks-grok-to-raise-200m-from-a16z-report/

Comments URL: https://news.ycombinator.com/item?id=42986667

Points: 2

# Comments: 2

The FAA’s Hiring Scandal

firebaze — Wed, 05 Feb 2025 05:25:19 +0000

Article URL: https://www.tracingwoodgrains.com/p/the-full-story-of-the-faas-hiring

Comments URL: https://news.ycombinator.com/item?id=42944203

Points: 739

# Comments: 666

Backpage.com

firebaze — Fri, 17 Jan 2025 12:54:26 +0000

Article URL: https://en.wikipedia.org/wiki/Backpage

Comments URL: https://news.ycombinator.com/item?id=42737018

Points: 1

# Comments: 0

New comment by firebaze in "Stimulation Clicker"

firebaze — Mon, 06 Jan 2025 22:22:40 +0000

This is so German

New comment by firebaze in "Ask HN: What are local LLMs good for?"

firebaze — Sat, 04 Jan 2025 20:54:11 +0000

We made an adapter (a specific CLI interface) for the LLM to interface with the app. Kind of like an integration test, just a little bit more sophisticated.

The LLM gets a prompt with the CLI commands it may use, and its "personality", and then it does what it does.

On the hardware-side, I personally have 2x 3090 cards on an AMD TR 79x platform with 128GB RAM, which yields around 12 token/sec for LLama 3.3 or Qwen 2.5 72B (Q5_k_m), which is okay (ingestion speed is approx double that)

If you want to know more details, feel free to drop me a message (username at liku dot social)

AWS Data Transfer Charges: Ingress Is Free (2022)

firebaze — Sat, 04 Jan 2025 06:22:03 +0000

Article URL: https://www.lastweekinaws.com/blog/aws-data-transfer-charges-ingress-actually-is-free/

Comments URL: https://news.ycombinator.com/item?id=42592876

Points: 2

# Comments: 0

New comment by firebaze in "Ask HN: What are local LLMs good for?"

firebaze — Fri, 03 Jan 2025 22:17:46 +0000

I'm using local LLMs for battle-testing an app by impersonating various personae with bad intentions (mostly on the social side, but also ones trying to break the system). Off-the-shelf cloud LLMs refuse such things (probably rightfully so); Llama 3.3 in q8, and qwen 2.5 72b even more so, are surprisingly good for this.

Also, even not-so-capable, smallish LLMs are able to be really good testers even outside the bad persona domain (given a good CLI interface). As long as your energy cost is okay-ish and you already have the hardware, that's quite a good use.

New comment by firebaze in "Notes on the New Deepseek v3"

firebaze — Thu, 02 Jan 2025 18:48:01 +0000

Ask a LLM to spell "Strawberry" one character per line. Claude's output, for example:

> Here's "strawberry" spelled out one character per line: s t r a w b e r r y

Most LLMs can handle that perfectly. Meaning, they can abstract over tokens into individual characters. Yet, most lack the ability to perform that multi-level inference to count individual 'r's.

From this perspective, I think it's the opposite. Something like the strawberry-tests is a good indicator how far the LLM is able to connect individually easy, but not readily interconnected steps.