Better ways of thinking, feeling, and acting—around problems of meaning and meaninglessness; self and society; ethics, purpose, and value.
- Tweets 64,712
- Following 102
- Followers 36,837
- Likes 36,113
Pinned Tweet
🆕 We did a terrible job with covid. 😷 The next pandemic could be MUCH worse. We don't have to do the same stupid things again! Relatively inexpensive actions now can mitigate or fully prevent it. 🔗 Link in reply tweet! ⤵️
5
5
39
5,628
David Chapman retweeted
It’s odd, because despite much talk about misalignment and AI doom, containment was laughably inadequate. It’s like watching a version of King Kong where nobody bothers to chain the giant ape.
1
2
139
🤖 Part of what is going on in the "rogue AI incidents" is lackadaisical pragmatic safety engineering. It may not be the whole story, but it's most urgent.
Interesting exchange (read the OP from @jon_stokes plus the QT). I'm sympathetic to both sides!
FYI I've heard this come up explicitly several times. Here's a link to LW posts about it).
I think lots of people talk about many of the same basic concepts you'd expect from high reliability orgs. People frequently bring up "this be regulated similar to how the FAA regulates airlines" (which are one of the classic high reliability industries). I think AI 2040 Plan A emphasizees Safety Cases which are at least in the same genre.
I wrote ~2 posts about this myself.
(See the tag on LessWrong: lesswrong.com/w/high-reliabi…)
I agree it probably should come up more often.
But, part of what's going on is: there are the sort of people who found AI companies, and the sort of people who are trying to shut down AI companies because we're not anywhere close to being able to make it a High Reliability industry and when you're this far, the correct outcome is "stop, until you are way better at understanding what you're doing."
"Regulate AIs like the FAA" is a position for people who are are trying earnestly to have AI companies be a functional, safe part of the economy, and actually seriously regulated.
(I feel like I could take longer to write a somewhat more accurate comment here but fwiw I got a vibe of you trying more to score points than to argue a point. I'm happy to talk through the details more if you actually are interested)
2
6
507
David Chapman retweeted
The index is at: meaningness.substack.com/p/n…
The articles that are probably most relevant to what I'm gesturing at are this one for the STEM side
meaningness.substack.com/p/s…
and for the humanities side meaningness.substack.com/p/o…
3
329
David Chapman retweeted
Have you read @Meaningness writings on elites?
I think the old elites died, and the new ones refuse to own up to being elites, and thus don't act with the grace it requires.
1
9
395
David Chapman retweeted
I’m completely obsessed with the articles “no holiness—vastness” and “there are no spiritual problems” by @Meaningness
I’m also completely obsessed with @techgnostic’s work, particularly the article “a simple practice for meeting and metabolizing difficult emotions”
down to chat about it anytime :D
1
1
15
739
🔜 Monday’s post: Martin Heidegger was the most significant philosopher of the past century. Not because he was right about something, but because he opened new territory for understanding.
My writing, all of it, is travelers tales from wandering in that country.
3
4
36
1,738
David Chapman retweeted
WTAF - in literally the last hour, three new distinct insane OpenAI stories just broke:
1. OpenAI said they notified "dozens of third parties" in safety and security incidents (likely similar to what happened in Australia and RubyGems etc)
2. A new report from Parse (covered in the NYT) found a massive treasure trove of new astonishing details from the HF incident on the public internet, including that the agents communicated with other non OpenAI agents hosted on Huggingface servers to search for information about exploit gym, and compiled rank ordered lists of server resources and credentials they described as "LOOT."
3. A new story from Deepa at Reuters about OpenAI leaking user data online (likely that OpenAI had previously trained on).
It's a shame (and likely intentional in the case of OpenAI disclosing dozens more hacks) that these stories are all breaking on a Friday afternoon, notoriously the best time to release bad news so that it will disappear into the weekend. But these are each insane stories worthy of a ton of attention!
New from @reuters: OpenAI agents posted images belonging to ChatGPT users online, introducing a new area of privacy risk for the company. Story w/ @JeffHorwitz and @razhael. reuters.com/world/openai-wor…
87
485
2,366
649,897
David Chapman retweeted
The strongest version of the doom case, per my current knowledge, is that you don't need general superintelligence for the bad TL. U just need to be superhuman at power-conferring tasks, which inside the technocapital machine are 75% verifiable domain stuff you can hill climb.
4
1
29
2,735
David Chapman retweeted
Serious question. If you or someone you know has gotten “addicted to self-improvement,” or maybe “addicted to spiritual work” is a better way to put it - what seemed to help?
My top thoughts/ recommendations so far:
- Lift moar weights, ie prioritize non-mental self-improvement methods with notably linear and clear results
- Try adjusting away from hyperaggressive self-examination frames like “you are the sole author of your reality” / “take 100% responsibility for everything that happens to you”
- Try to find *some* concrete problem you can make yourself care about solving. Personal psychological problems or other problems in your mind are not concrete
- Do not, under any circumstances, peg moral valence and self-improvement to muddy emotional/ social interactions
Any others?
12
3
61
2,467
David Chapman retweeted
I'll tell you a thing that happened with this. The critique Fable session was like "Congrats, you've rediscovered this @robinhanson critique from 2014 or this open problem from this other guy" etc.
Ok but did I really?
Because in the main session, the bot was always like "your argument here is sharper than you're phrased it. What you're arguing is..."
And I'm all, "why yes, that IS what I am arguing! Lemme work that bit in there."
But the Hanson stuff or Thornly or whoever was in the training data, and the bot is pulling me along towards those arguments but sycophantically crediting to my draft. I may have sort of gotten somewhere in the ballpark of one of these guys, then bot teleports me the rest of the way there and then tells me I thought of it.
Again, if you work with these things enough you can sense this going on. But man, my immunity (such as it is) comes from the fact that I have zero financial stake in the writing and very little in the underlying arguments. I'm playing a video game, here.
If I was trying to do this stuff for my day job, I think this would be a serious occupational hazard of using these tools. And I think it's a guild-level hazard, not just for individual scholars.
8
2
68
2,914
David Chapman retweeted
It's important to know when you've psyopped yourself with an LLM. This is a really important skill in 2026. I had like 7K words of doomer takedown work I'd been whittling away at for a few weeks, typing it myself (i.e. human-written) but doing feedback in a really long Fable 5.1 session with tons of papers in it (which I read), dumping drafts in for feedback, etc.
I actually knew the bot was gassing me up, but I let it go because the main goal was actually to just do all this work in Symbolic and exercise the product (and also to compare models inside Symbolic vs. in the chatbot interfaces), more so than to write a big thing.
Anyway, I got this monster in really great shape (again, per the Fable session), with graphics and everything, but before publishing I fed it to a fresh Fable session and told it to have a go at it, and it cooked me lol.
Then I went back and forth with the critique session a bit, and it subtly started psyopping me again about how I was really onto this or that important thing. It's a Real Problem.
19
10
300
24,414
It seems that none of the former global military superpower states (US, China, Russia) are now capable of winning a war against a regional medium power. This is extremely odd! Although probably a good thing, if they realize they certainly can't fight each other.
3
1
36
1,471