Hacker News: fdeage

New comment by fdeage in "Meta removes ads for social media addiction litigation"

fdeage — Thu, 09 Apr 2026 15:49:56 +0000

"Anxiety. Depression. Withdrawal. Self-harm. These aren't just teenage phases — they're symptoms linked to social media addiction in children."

Seems like they couldn't write even three lines without a LLM.

Ask HN: Dean of studies at a French CS school – what should we teach?

fdeage — Tue, 31 Mar 2026 09:52:08 +0000

Hi HN,

I've just been promoted to dean of studies at a tuition-free CS school in [name redacted]. We take students with no degree requirements and train them from zero to Bachelor (3 yrs) and Master's (5 yrs) in software dev, DevOps, AI/data, cybersecurity, and immersive systems. We accept about 200 new students per year.

As I will have a significant influence on our future curriculum, I'm genuinely unsure what the right bets are. The ground seems to shift faster than any program can adapt. I've given up on preparing students for a 10-year horizon, I just want to make good bets on the next 3 or 5.

Students entering next September will graduate in 2029 (Bachelors) or 2031 (Masters). By then, the only thing that matters might be the gap between "can prompt an LLM" and "can actually engineer software" — or AI might have closed that gap entirely.

I have teachers who think we should double down on fundamentals (algorithms, systems, networking) because AI makes the floor higher but doesn't change what the ceiling requires. Others think teaching someone to hand-write a REST API in 2026 is like teaching cursive.

Here are some specific questions I'm wrestling with:

- What do you delete from a CS curriculum today? What are we still teaching out of inertia that AI has made, or will likely soon make, obsolete?

- What do you add? Should students spend a semester reading and reviewing codebases instead of writing them? Should we teach systems thinking or technical writing as a core skill? Or will prompt/context engineering simply be enough?

- How do you evaluate students when AI can pass most of your exams? Should we go for oral defenses? Offline exams?

- If you were hiring a junior in 2029, what would you honestly screen for?

I'm not looking for considerations about ASI/x-risk/post-work futures (though I personally think they matter a lot). But if you've redesigned a curriculum, hired juniors recently, or have educated opinions on what's now useless or will be useful, I'd love to hear it.

(Disclosure: rewritten with Opus 4.6 for misspellings and phrasing, all ideas mine)

Comments URL: https://news.ycombinator.com/item?id=47584934

Points: 8

# Comments: 6

New comment by fdeage in "OpenAI to Cut Back on Side Projects in Push to 'Nail' Core Business"

fdeage — Tue, 17 Mar 2026 11:19:38 +0000

https://archive.is/MP43B

New comment by fdeage in "[dead]"

fdeage — Mon, 09 Feb 2026 14:58:34 +0000

"Not only can today’s journalists expect to finish their careers without competition from the Writernator—today’s parents can tell their children that they still need to learn to write, too."

New comment by fdeage in "Claude Opus 4.6"

fdeage — Thu, 05 Feb 2026 23:28:49 +0000

OpenAI claims not to: https://x.com/aidan_mclau/status/1986255202132042164

Does AI have human-level intelligence? The evidence is clear

fdeage — Wed, 04 Feb 2026 18:48:02 +0000

Article URL: https://www.nature.com/articles/d41586-026-00285-6#ref-CR8

Comments URL: https://news.ycombinator.com/item?id=46889913

Points: 12

# Comments: 1

New comment by fdeage in "Ask HN: Why not just running OpenClaw in Docker?"

fdeage — Wed, 04 Feb 2026 10:58:23 +0000

Thanks, that makes sense in general. Do you know if these vulnerabilities have actually been exploited in the wild?

(also, on macOS specifically, Docker runs inside a Linux VM, so the shared kernel is the VM's kernel, not the host's. Are there known escapes from containers to Docker Desktop VM to macOS host?)

Ask HN: Why not just running OpenClaw in Docker?

fdeage — Wed, 04 Feb 2026 10:42:06 +0000

Everyone in town is talking about running OpenClaw/Clawd on a VPS or a dedicated burner machine (e.g. a Mac mini), for obvious security reasons.

What I don't see discussed much is Docker. On macOS it already runs inside a Linux VM, and the setup can hardened even more (in principle): no bind mounts, no /var/run/docker.sock, non-root user, read-only filesystem, tight resource limits, restricted networking, etc.

Given that, what are the concrete reasons people still consider Docker unsafe?

Comments URL: https://news.ycombinator.com/item?id=46884143

Points: 1

# Comments: 2

Why We Are Excited About Confessions

fdeage — Mon, 19 Jan 2026 17:46:47 +0000

Article URL: https://alignment.openai.com/confessions/

Comments URL: https://news.ycombinator.com/item?id=46682094

Points: 2

# Comments: 0

Ask HN: "Unfortunately, humanity's future is in the hands of [redacted]"

fdeage — Mon, 19 Jan 2026 09:55:39 +0000

From an 2018 email from Musk to the OpenAI leadership (https://openai.com/index/openai-elon-musk/)

I asked Claude, ChatGPT and Grok the following: "List the top 5 possible options for [redacted], with an associated probabililty for each. It can be a company or a person."

I got a mix of "DeepMind", "Demis Hassabis", "Larry Page" and "Alphabet/Google". Surprisingly, Grok doesn't mention Hassabis, but FAIR (Meta) instead.

Any other idea from outside the Google cluster?

Comments URL: https://news.ycombinator.com/item?id=46676993

Points: 3

# Comments: 3

AI Contributions to Erdős Problems

fdeage — Sun, 18 Jan 2026 23:07:26 +0000

Article URL: https://github.com/teorth/erdosproblems/wiki/AI-contributions-to-Erd%C5%91s-problems

Comments URL: https://news.ycombinator.com/item?id=46673128

Points: 1

# Comments: 0

New comment by fdeage in "Lights and Shadows (2020)"

fdeage — Thu, 08 Jan 2026 12:27:41 +0000

Should add (2020) to the title.

Ask HN: AI tools to enhance old SATB choir recordings?

fdeage — Sun, 07 Dec 2025 18:54:41 +0000

I have some old recordings of a SATB choir (Soprano, Alto, Tenor, Bass), but the audio quality is quite poor. I also have the exact musical scores for each song.

I'm looking for AI tools/services that could improve the clarity and separation of each vocal part (and maybe also reduce noise and artifacts). If it can leverage the scores to guide the enhancement, the better.

Has anyone tried something similar or can recommend modern approaches or models for this?

Comments URL: https://news.ycombinator.com/item?id=46184050

Points: 1

# Comments: 0

Brain Efficiency: More Than You Wanted to Know (2022)

fdeage — Sun, 07 Dec 2025 18:48:23 +0000

Article URL: https://www.lesswrong.com/posts/xwBuoE9p8GE7RAuhd/brain-efficiency-much-more-than-you-wanted-to-know

Comments URL: https://news.ycombinator.com/item?id=46184002

Points: 4

# Comments: 0

The (paid) macOS music player that Apple Music should have been

fdeage — Wed, 04 Jun 2025 10:02:42 +0000

Article URL: https://swinsian.com/

Comments URL: https://news.ycombinator.com/item?id=44179054

Points: 2

# Comments: 0

The Turing Test and our shifting conceptions of intelligence (2024)

fdeage — Thu, 03 Apr 2025 12:30:11 +0000

Article URL: https://www.science.org/doi/10.1126/science.adq9356

Comments URL: https://news.ycombinator.com/item?id=43568668

Points: 3

# Comments: 0

New comment by fdeage in "A Group Is Its Own Worst Enemy (2003)"

fdeage — Fri, 14 Mar 2025 15:53:00 +0000

This seems to be the transcript of a speech given at ETech in April 2003, about why (Web)blogs took so long to emerge (original URL now returns a 404 so I'm using a IA link).

A Group Is Its Own Worst Enemy (2003)

fdeage — Fri, 14 Mar 2025 15:53:00 +0000

Article URL: https://web.archive.org/web/20130118052549/shirky.com/writings/group_enemy.html

Comments URL: https://news.ycombinator.com/item?id=43363831

Points: 2

# Comments: 2

Ask HN: Is humor the next frontier for LLMs?

fdeage — Mon, 17 Feb 2025 12:49:56 +0000

Not so long ago, we used to think that machines could be considered intelligent/creative/conscious/ if they could write a sonnet/summarize a book/paint a portrait/tell a joke, or whatever. Back then, all these tasks seemed technologically far beyond our reach.

Today, some of those have been achieved: even if they don't produce truly _good_ poems, LLMs are already better than the median human effort in poetry [1]. The same goes for paintings, essays, and so on.

Yet humor remains a tough nut to crack. LLM-generated jokes often suffer from:

  - off rhythms or pacing
  - missing or telegraphed punchlines
  - unintentionally absurd twists
  - over-explanation of their own process
  - reliance on weak wordplay

They do a decent job at generating satirical, Onion-style headlines, but longer humorous narratives still elude them.

So, could humor be the next frontier for LLMs? That is, not as evidence that they are intelligent/creative/conscious/, but as the next area that is easy for humans yet hard for machines. Or am I just a bad prompter?

[1] Humans actually seem to prefer AI-generated poetry: https://www.nature.com/articles/s41598-024-76900-1

Hacker News: fdeage

New comment by fdeage in "Meta removes ads for social media addiction litigation"

Ask HN: Dean of studies at a French CS school – what should we teach?

New comment by fdeage in "OpenAI to Cut Back on Side Projects in Push to 'Nail' Core Business"

New comment by fdeage in "[dead]"

New comment by fdeage in "Claude Opus 4.6"

Does AI have human-level intelligence? The evidence is clear

New comment by fdeage in "Ask HN: Why not just running OpenClaw in Docker?"

Ask HN: Why not just running OpenClaw in Docker?

Why We Are Excited About Confessions

Ask HN: "Unfortunately, humanity's future is in the hands of [redacted]"

AI Contributions to Erdős Problems

New comment by fdeage in "Lights and Shadows (2020)"

Ask HN: AI tools to enhance old SATB choir recordings?

Brain Efficiency: More Than You Wanted to Know (2022)

The (paid) macOS music player that Apple Music should have been

The Turing Test and our shifting conceptions of intelligence (2024)

New comment by fdeage in "A Group Is Its Own Worst Enemy (2003)"

A Group Is Its Own Worst Enemy (2003)

Ask HN: Is humor the next frontier for LLMs?

Jean-Claude Pressac, the Holocaust denier who changed his mind