Hacker News: tomrod

New comment by tomrod in "Accelerating GPT-5.6 Sol Ultrafast"

tomrod — Thu, 13 Aug 2026 20:27:28 +0000

ChatJimmy is a much smaller model and, AFAIK, has no reasoning capability. Absolutely insane raw speed, like a supercar, while Sol is more like a freight truck.

New comment by tomrod in "Grok Bot"

tomrod — Tue, 11 Aug 2026 23:39:30 +0000

AI Loopidity in other words

New comment by tomrod in "Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots"

tomrod — Tue, 11 Aug 2026 04:28:00 +0000

Oh I agree that's not unreasonable. I wonder about the harnessing heft required to make it feasible though. If I say I'd like it a bit warmer, a tool can deterministically bump a few degrees while an LLM might bump it 10C. So do we limit the tool's range?

New comment by tomrod in "Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots"

tomrod — Tue, 11 Aug 2026 02:51:23 +0000

I confused by the dispatch model. Tool calls typically need some reasonability to be deterministic and, more importantly, predictable in response (o/w GIGO). Why would the thermostat need to interpret a voice command that the node capturing the voice command would not interpret instead?

New comment by tomrod in "Illinois just passed a law that puts Linux on the hook for age verification"

tomrod — Mon, 10 Aug 2026 20:50:06 +0000

I'm surprised we're not seeing heaps of lawsuits here. Age verification in general violates privacy.

New comment by tomrod in "Tailscale didn't stop the Hugging Face intrusion"

tomrod — Fri, 31 Jul 2026 19:52:40 +0000

Just so you know, Tailscale tech has always impressed me, but your responsiveness here has cemented my appreciation of your org and I look forward to bringing yall in as our company hits the growth phase.

May y'all continue to be customer focused.

New comment by tomrod in "Show HN: CheapFoodMap – A map of good meals under $10"

tomrod — Thu, 30 Jul 2026 10:30:23 +0000

FoodGaloreMap. Focuses on the abundance without the negative miser association. Or FoodMiser works too, just lean into it as a beautiful brand.

Because good and plentiful food for cheap is abundance, and good food for relatively cheap is miser's dream.

New comment by tomrod in "Kimi K3 Architecture Overview and Notes"

tomrod — Wed, 29 Jul 2026 17:16:32 +0000

Thoughtleader as a brand is oversubscribed. Trusted is way less so -- and he is both.

New comment by tomrod in "Kimi K3 Architecture Overview and Notes"

tomrod — Wed, 29 Jul 2026 17:16:07 +0000

His was the only Packt book (on ML, circa 2019 IIRC) that I have never regretted purchasing. Clear thinker, attention to details, and clean coder. Highly recommend.

New comment by tomrod in "The New AI Superpowers: Focus and Followthrough"

tomrod — Mon, 27 Jul 2026 18:37:33 +0000

Brilliant irony!

New comment by tomrod in "The New AI Superpowers: Focus and Followthrough"

tomrod — Sun, 26 Jul 2026 19:44:45 +0000

Steve Jobs said something about good and great artists.

New comment by tomrod in "Kain – A non-von Neumann language (Zero GC, Python syntax, no borrow checker)"

tomrod — Sat, 25 Jul 2026 23:44:53 +0000

Regarding me being a pretentious nimrod: no doubt there are many subscribers to this frame of mind.

Clarifying my structural comment: since LLMs are next token generators, they fall into the "aren't even wrong" side of lying versus truth-telling. Truth reflects a reality to which LLMs are not directly privy. Hallucinations are baked into the source. Harnesses on agents help to ground some, but they don't fully solve the issue. This doesn't make them useless. Just unreliable narrators (i.e. liars).

New comment by tomrod in "The new rules of context engineering for Claude 5 generation models"

tomrod — Sat, 25 Jul 2026 21:50:52 +0000

This made me belly laugh.

New comment by tomrod in "Kain – A non-von Neumann language (Zero GC, Python syntax, no borrow checker)"

tomrod — Sat, 25 Jul 2026 19:36:57 +0000

Always, structurally.

New comment by tomrod in "Kain – A non-von Neumann language (Zero GC, Python syntax, no borrow checker)"

tomrod — Sat, 25 Jul 2026 17:18:01 +0000

These benchmarks look a bit off - many times faster than C++ / Rust? Are these independently implemented?

New comment by tomrod in "Three out of four job postings won't tell you the pay"

tomrod — Sat, 25 Jul 2026 17:12:54 +0000

Yuck.

Post the pay band if you want to be taken seriously.

New comment by tomrod in "Silicon Valley Is Completely Divided over Chinese AI"

tomrod — Sat, 25 Jul 2026 17:11:50 +0000

Paywalled.

"Chinese AI" models (or, not corporate models serving REST endpoints with highly fluctuating prices) are fine if you can afford to self-host or otherwise work independently.

All things are systems – the Idlewild Acres case

tomrod — Sat, 25 Jul 2026 17:10:18 +0000

Article URL: https://www.roderick.dev/writing/2026-07-25-systems-axioms-impacts/

Comments URL: https://news.ycombinator.com/item?id=49049403

Points: 2

# Comments: 0

New comment by tomrod in "Claude Cookbook"

tomrod — Fri, 24 Jul 2026 11:27:13 +0000

That sounds like a great idea. Let's stop killing people altogether. It's almost entirely unnecessary.

New comment by tomrod in "The arguments against open source AI are bad"

tomrod — Thu, 23 Jul 2026 19:18:56 +0000

Audit them for what, exactly?

Backdoor hardware is straightforward -- its letting people in that the customer doesn't authorize.

Open weights means you aren't tied to any harnessing. So the security domain to be really concerned about is limited to weights only, and I welcome pushback on that.

Are we thinking adversarial injection? Lying about history / propaganda infusion? What is the angle that makes a non-sovereign model dangerous in a way a sovereign model isn't?