frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Intelligence Cannot Be Trained?

4•hyperzzw•8h ago
What do you think?

Comments

uncanny2•7h ago
Wut?

Intelligence is the mitigation of uncertainty, if it isn’t mitigating uncertainty it is not intelligent.

All intelligence is “trained” at some level, even “genetic” or “compositional” intelligence is “trained.” Your understanding of what that is and how it works is lacking.

Training takes time and may be conflated or inaccurate or inadequate, yet “training” (iterative preparation, usually with synthesized samples or circumstances) is necessary.

You remind me of a friend who once said he tried working out but stopped after he (inexperienced) worked out real hard and then hurt for two weeks. Working out didn’t work for him. Is training not working for you?

If you haven’t noticed, “IQ” testing has to be updated from time to time as the “average” creeps upward. 100 is supposed to always be the average, this is adjusted over generations as whatever qualities these sample are rising over time.

ahmed-fathi•3h ago
Intelligence is a high tolerance for not knowing yet
havealight•1h ago
The most mind-blowing thing LLMs proved to me was that knowledge is only prediction at scale.

If you see “The quick brown fox jumped over the lazy dog” 80,000 times - and enough other sentences involving all those words to see how they generally fit together - then the system KNOWS what jumped over the lazy dog. It “knows” that it was none other than the quick brown fox.

Any system that models things (saves the relations) in a way that makes it easier and easier to output a prediction about it each time is developing “knowledge” of it, is “learning” about it, and if it can reliably articulate abstract truths about it, is “intelligent” in that particular way.

When I say somebody knows C better than anyone else, I mean they can reliably articulate all the right things in the right order - they “know” it, they have intelligence, they’ve modeled it enough to make their own reproductions, whether partial or derivative and so-on.

Being consciously aware of it might be an entirely separate system.

Training mRNA Language Models Across 25 Species for $165

95•maziyar•3d ago•27 comments

Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw

1006•firloop•23h ago•766 comments

Ask HN: How Do You Relax?

5•azeirah•2h ago•2 comments

Intelligence Cannot Be Trained?

4•hyperzzw•8h ago•3 comments

Ask HN: Who is hiring? (April 2026)

274•whoishiring•3d ago•346 comments

Ask HN: Who wants to be hired? (April 2026)

85•whoishiring•3d ago•336 comments

Anthropic to limit Using third-party harnesses with Claude subscriptions

17•guiyuwei•23h ago•5 comments

Third-party Claude harnesses will now draw from extra usage

12•iBelieve•23h ago•4 comments

Ask HN: Distributed data centers in our basements

73•cmos•4d ago•65 comments

Ask HN: Learning resources for building AI agents?

5•7e10•13h ago•5 comments

Tell HN: YouTube history setting privacy implications

5•tinaclaussen•14h ago•1 comments

Ask HN: Why still embed heavy 3rd-party iFrames for simple social proof?

4•LordKode•17h ago•0 comments

Ask HN: Should we collectively stop spell checking and fixing grammar

5•sankalpnarula•1d ago•15 comments

Ask HN: Analysis of credit card receipts to show one does not buy alcohol

4•bryanrasmussen•20h ago•10 comments

Tell HN: Chrome says "suspicious download" when trying to download yt-dlp

308•joering2•4d ago•97 comments

Lightweight IDE to Pair with Claude Code?

4•zupancik•23h ago•1 comments

Ask HN: Are agents _still_ writing most of your code?

5•etamponi•1d ago•4 comments

What Is a Collation, and Why Is My Data Corrupt? – PG Phridays with Shaun Thomas

3•pgedge_postgres•1d ago•0 comments

Ask HN: What dev tools do you rely on that nobody talks about?

42•crcsmnky•3d ago•36 comments

Ask HN: What is your dev set up like?

13•break_the_bank•2d ago•27 comments

Stripe closed my UAE business account and is withholding $3.5K

25•alganzory•2d ago•5 comments

Ask HN: Thought Exercise regarding RX dependency

3•Bender•1d ago•0 comments

Ask HN: How relevant is low code according to you in today’s world with Claude?

8•kinj28•2d ago•7 comments

Ask HN: Advice on Solo Launching

5•acaciabengo•1d ago•2 comments

Ask HN: Academic study on AI's impact on software development – want to join?

31•research2026•4d ago•15 comments

Ask HN: European Tech Alternatives?

103•BrunoBernardino•2d ago•54 comments

Ask HN: What are you moving on to now that Claude Code is so rate limited?

7•esperent•1d ago•14 comments

Ask HN: Whether there is LLM or other strong NLP behind Hacker News?

3•Jet_Xu•1d ago•1 comments

Fixed a llama.cpp bug silently disabling Vulkan GPU on all 32-bit ARM devices

3•perinban•1d ago•0 comments

Ask HN: Is there any interest in a native Qt/C++ Discord client?

5•txtsd•2d ago•2 comments