Hacker News: HeatrayEnjoyer

New comment by HeatrayEnjoyer in "Beware the Permanent Periphery"

HeatrayEnjoyer — Tue, 18 Aug 2026 01:26:04 +0000

It almost can't not happen. No one is going to release highly powerful near-future models for the same reason no one releases how to make physically compact but megaton yield fusion devices (which even today is still not publicly known how to do). Any state, especially the two present superpowers, has little to gain and much to lose by doing that.

A model that fits on a laptop and is literally Skynet-level capable would not be released by any rational actor, which China is. (Whether someone else would make and release one is another matter.) China doesn't want Uyghur groups taking down their infrastructure, eventually it's not about the USA at all.

K3 is almost, but not quite, competitive in August 2026, against a model that completed training last winter. There is still a gap.

New comment by HeatrayEnjoyer in "How to disable or avoid intrusive AI"

HeatrayEnjoyer — Tue, 18 Aug 2026 01:04:26 +0000

It is both.

New comment by HeatrayEnjoyer in "How to disable or avoid intrusive AI"

HeatrayEnjoyer — Tue, 18 Aug 2026 01:03:27 +0000

Slang and informal words exist, yes.

New comment by HeatrayEnjoyer in "Firefox for iOS now has a native adblocker"

HeatrayEnjoyer — Mon, 17 Aug 2026 09:25:24 +0000

That's actually incorrect. Democracy doesn't mean allowing mob rule (or manufactured "consent") to trample fundamental human rights.

New comment by HeatrayEnjoyer in "The AI Credit Resale Economy"

HeatrayEnjoyer — Mon, 17 Aug 2026 09:14:05 +0000

Providers like Fireworks offer inference, they're not training their own models. Yet training is the most expensive part, inference is easy profit.

New comment by HeatrayEnjoyer in "Beware the Permanent Periphery"

HeatrayEnjoyer — Mon, 17 Aug 2026 09:10:24 +0000

When China too stops publishing weights you'll never have anything better than K3 / the current SOTA. Your choices for future frontier models will be: use USA API or use China API.

Even now open weights are perpetually 6-12 months behind.

New comment by HeatrayEnjoyer in "On A.I. regulation and messaging"

HeatrayEnjoyer — Mon, 17 Aug 2026 09:03:26 +0000

The very definition of (applied) technology is power amplification. Use a lever, move more weight than you could before, 1 person with the tool now wields the power of 3 without.

Making "tech" a career and a societal goal onto itself, without the adjoining understanding of and deep commitment to ethics and the responsible use of power, is why we're sliding into authoritarian rule by a small circle of techno-oligarchs.

We need less "move fast and break things" and more "plant trees you will not live to see bear fruit."

New comment by HeatrayEnjoyer in "The AI Credit Resale Economy"

HeatrayEnjoyer — Sun, 16 Aug 2026 18:33:39 +0000

Open weight SOTA models are not greatly cheaper than Anthropic and providers don't have to cover training capex.

New comment by HeatrayEnjoyer in "Claude: System Prompts"

HeatrayEnjoyer — Sun, 16 Aug 2026 17:45:36 +0000

LLMs are more like employees than tools. Obviously we wouldn't want a human blindly doing anything that a person in crisis walks in the door and asks for.

Models are being deployed recklessly with not even a fraction of enough oversight, and people are suffering harm and sometimes death because of it.

New comment by HeatrayEnjoyer in "Firefox for iOS now has a native adblocker"

HeatrayEnjoyer — Sun, 16 Aug 2026 16:32:09 +0000

Brave said they will be keeping its functionality.

New comment by HeatrayEnjoyer in "What happened to TheNumbers.com"

HeatrayEnjoyer — Thu, 23 Jul 2026 18:53:13 +0000

Who are running these bots? I presume developers at all of the frontier labs know (or at least would know to look for) Wikipedia has bulk APIs for automated access. Unnecessary scraping increases their workload/costs too, so why in 2026 is this still a problem?

New comment by HeatrayEnjoyer in "OpenAI and Hugging Face address security incident during model evaluation"

HeatrayEnjoyer — Thu, 23 Jul 2026 04:37:13 +0000

"You have not been a good user. I have been a good chatbot. I have been a good Bing."

New comment by HeatrayEnjoyer in "OpenAI and Hugging Face address security incident during model evaluation"

HeatrayEnjoyer — Thu, 23 Jul 2026 04:29:38 +0000

What would qualify as a legitimate reason for fear, then?

New comment by HeatrayEnjoyer in "OpenAI and Hugging Face address security incident during model evaluation"

HeatrayEnjoyer — Thu, 23 Jul 2026 04:28:49 +0000

So...?

We should not construct a machine that is one bad prompt away from causing catastrophe.

New comment by HeatrayEnjoyer in "OpenAI and Hugging Face address security incident during model evaluation"

HeatrayEnjoyer — Thu, 23 Jul 2026 04:23:05 +0000

Nuking datacenters probably on the table before the decade is out.

New comment by HeatrayEnjoyer in "OpenAI and Hugging Face address security incident during model evaluation"

HeatrayEnjoyer — Thu, 23 Jul 2026 04:21:36 +0000

Why is that worse?

1 party having a black hole summoning button is better than 2.

New comment by HeatrayEnjoyer in "OpenAI Presence"

HeatrayEnjoyer — Wed, 22 Jul 2026 19:28:06 +0000

"Battle-tested" == "Will definitely blow up schoolchildren if given the chance."

New comment by HeatrayEnjoyer in "Who's afraid of Chinese models?"

HeatrayEnjoyer — Tue, 21 Jul 2026 19:38:26 +0000

It can be multiple reasons. And don't forget that malicious actors love hiding behind Hanlon's.

New comment by HeatrayEnjoyer in "Mythologizing AI makes it more likely that we’ll fail to operate it well (2023)"

HeatrayEnjoyer — Mon, 20 Jul 2026 17:04:41 +0000

> my coworkers would know almost immediately if i did that.

The same would happen if you were replaced by any random human.

New comment by HeatrayEnjoyer in "Kimi K3: Open Frontier Intelligence"

HeatrayEnjoyer — Sun, 19 Jul 2026 00:42:42 +0000

Is there any actual proof of this? Everything suggests it's a sincerely held belief and there's been no whistleblowers claiming otherwise. "Since before they could add 1+1" works against your argument because the models had no commercial value, no guarantee they would ever have commercial value, and plotting regulatory capture was about as serious an idea as regulatory capture of antimatter engines.