Hacker News: flimflamm

New comment by flimflamm in "Croatia declared free of landmines after 31 years"

flimflamm — Sat, 28 Feb 2026 09:45:10 +0000

Cost/manufacturing complexity. If you are country struggling to defend your self you don't think problems in 30 years if today problem is does the country exists or not. Might be difficult to put your self to a small defending countries shoes which is absolute running our of resources.

New comment by flimflamm in "Claws are now a new layer on top of LLM agents"

flimflamm — Sun, 22 Feb 2026 09:54:18 +0000

If you don't have workflows which repeat in inet you don't need openClaw. - Messages from school where to react - Getting payments from someone and tracking that you get them - Summary of news the way you like it from sources you like it every day - Integrated task lists reminders - Drafting taxation reports based on spending etc etc.

New comment by flimflamm in "Claws are now a new layer on top of LLM agents"

flimflamm — Sun, 22 Feb 2026 09:41:46 +0000

I just don't trust "the claw" so I build following system - Docker 1: * Locked up Claw docker - user level priv. Access outside to "one port" only.

- Docker 2: * Tool gateway with pre-baked commands - openclaw can only index what command to execute * Keys are here * Telegram hook to approve all "post" commands i.e. sending email or posting something somewhere.

- Docker 3: * LLM gateway keeping track of cost and routing

New comment by flimflamm in "Sopro TTS: A 169M model with zero-shot voice cloning that runs on the CPU"

flimflamm — Fri, 09 Jan 2026 17:00:28 +0000

No as you could have multiple examples of expected output in single prompt. You should just "ask" - that's zero shot. If you "ask + provide" examples then you are in the n+1 shot realm. But I suppose terminology is shifting from this...

New comment by flimflamm in "Reasoning models reason well, until they don't"

flimflamm — Fri, 31 Oct 2025 10:54:39 +0000

What confused me is the fact that in the paper all logical steps are give. It basically check that when all relevant facts are provided explicitly as links , how far and how complex a chain can the model correctly follow before it breaks down?

So it's simpler than "reasoning". This is not necessarily a bad thing as it boils down the reasoning to a simpler, more controlled sub problem.

New comment by flimflamm in "How I turned Zig into my favorite language to write network programs in"

flimflamm — Mon, 27 Oct 2025 07:34:07 +0000

Is this just buttering the cost of switches by crippling the optimization options compiler have?

New comment by flimflamm in "A definition of AGI"

flimflamm — Mon, 27 Oct 2025 07:32:04 +0000

I would focus on lowest of the axis. It does not help if some of the axis are at 100% if one of the axis is lacking.

New comment by flimflamm in "How to make a Smith chart"

flimflamm — Sat, 25 Oct 2025 07:23:53 +0000

Really really nice. Thanks for this!

New comment by flimflamm in "IRS open sources its fact graph"

flimflamm — Thu, 16 Oct 2025 05:47:54 +0000

Are the rules versioned somehow? I didn't see that.

New comment by flimflamm in "DeepMind and OpenAI win gold at ICPC"

flimflamm — Thu, 18 Sep 2025 06:03:18 +0000

Good to note that OpenAI solved 12/12 and DeepMind 10/12.

New comment by flimflamm in "Microsoft doubles down on small modular reactors and fusion energy"

flimflamm — Tue, 09 Sep 2025 08:11:12 +0000

I wonder when Elon will go to nuclear business as green values have gone down (solar tiles anyone?)

New comment by flimflamm in "Mistral raises 1.7B€, partners with ASML"

flimflamm — Tue, 09 Sep 2025 08:05:07 +0000

I wonder what process in ASML is such that Mistral group would bring something new there...

New comment by flimflamm in "We hacked Burger King: How auth bypass led to drive-thru audio surveillance"

flimflamm — Sun, 07 Sep 2025 09:25:11 +0000

Depends on the country. In Finland, it's ok to record your own discussions. Whether the recorder is BK (a third party) or the cashier is an interesting question, though.

New comment by flimflamm in "We hacked Burger King: How auth bypass led to drive-thru audio surveillance"

flimflamm — Sun, 07 Sep 2025 09:23:52 +0000

Creating a database of recordings without user being able to know/influence is clearly violation of GDPR IF there is PII. That's going to be costly for BK.

New comment by flimflamm in "How attention sinks keep language models stable"

flimflamm — Sat, 09 Aug 2025 04:27:50 +0000

Hah they kind of found "NULL" pointer in LLMs.

New comment by flimflamm in "Valve confirms credit card companies pressured it to delist certain adult games"

flimflamm — Sat, 19 Jul 2025 06:55:01 +0000

Have you looked in to "PSD2 and Open Banking regulation in Europe" ?

New comment by flimflamm in "Bulgaria to join euro area on 1 January 2026"

flimflamm — Wed, 09 Jul 2025 04:35:56 +0000

Except if you are a Finnish person who lives right next to Sweden (SEK) and Norway (NOK) who are not using Euro (and Russia but that's a different story).

New comment by flimflamm in "How to Build Conscious Machines"

flimflamm — Sun, 15 Jun 2025 07:12:06 +0000

I wonder if 6 would be understanding own thinking. Currently humans don't understand this. Thoughts just pop in to our heads and we try to explain what caused them.

New comment by flimflamm in "AI code review: Should the author be the reviewer?"

flimflamm — Fri, 02 May 2025 07:04:35 +0000

Surely you can have the same LLM review the code. But treat output of that process similar to linters.

That one tool (LLM) didn't show issues - good. Then lets roll slaves back and start really checking if the meaning is correct and is the code implementing the right things the right way.

New comment by flimflamm in "Text-to-Visual with AI by Excalidraw"

flimflamm — Thu, 24 Apr 2025 13:36:31 +0000

Can't CTRL-Scroll to zoom to the generated image which is with so small text that I can't read it.