My substack: https://tomasbjartur.substack.com/
EPUB of all my fiction: https://tomasbjartur.com/The_Origami_Men_and_Other_Stories.epub
Fiction:
My Twitter: https://x.com/bjarturtomas
LLM policy: None of my posts or comments are written by, or in collaboration with, an LLM. This includes Beauty and the Beast. Occasionally, I will use an LLM for proofreading/feedback/research/sanity checking, but all words are my own.
I’m not involved in judging submissions.
I very much disagree with the implicit reference class you use when analyzing self-sovereign AIs, though the political economy frame is slightly more sane, in my reading, than the purely economic one. I imagine you’ve heard all the arguments from smarter people than I, but in case you have not:
Byrnes’ criticism of the econ frame: https://www.lesswrong.com/posts/xJWBofhLQjf3KmRgg/four-ways-learning-econ-makes-people-dumber-re-future-ai
Reasons I expect things to happen faster post ASI than even 2027 projects:
https://www.lesswrong.com/posts/Lc8v92pGPtjF3gsCv/some-quick-thoughts-on-ai-2027
Latent potential for discontinuous power grabs:
https://www.lesswrong.com/posts/uDAWPNwPJfEY7oroF/self-hosting
https://www.lesswrong.com/posts/JcavsPku6RR9hcujz/slightly-super-persuasion-will-do
I read this tweet as claiming this article was written by fable:
Is this true? If so, it needs to go in an LLM block.
Presumably, some of these websites were hacked and have standing to sue. Such an effort would be worth funding.
I created a sequence for all my fiction: https://www.lesswrong.com/s/uHstmwwxYmL9YDoZm
Fair enough. And perhaps I am misreading, but their mainline scenario does not assume a race to the smallest self replicator. If they are, I will take this post down. Otherwise, I think it has things worth saying. I would fix or add a mention of the dropdown, but I am writing a fiction post and don’t want to spend too much time on this.
After Lobsang’s Children, this is my favourite of my InkHaven posts.
I very much enjoyed reading this. It is deeply weird in a way useful to me.
The Commonwealth Short Story Prize (which is decently prestigious) was won by a story called The Serpent in the Grove.
>Did you feel anything from 200mg PX at all?
not really. i can’t write without caffeine/nicotine. and PX/nicotine didn’t work.
1 cup of coffee.
100 mg every morning.
I tried it for a few weeks. it does basically nothing for me even at 200 mgs.
Richard tests his audience to see if they can tell the difference between his writing and Claude’s. They mostly can’t.
It sure does look like it’s almost over for human-written essays—though Hanania’s probably right about some people getting grandfathered in. I am unsure if I will be once fiction falls. The age of the pseudonymous weirdo is almost over. Or at least, most pseudonymous weirdos who rise to prominence will be AIs—or be accused of being AIs. As a pseudonymous weirdo, this makes me sad. Though this is something I have been expecting, and I would not have written as much as I have in the last seven months if not for this expectation.
As far as I can tell, fiction-writing ability and humor are lacking in the best models. But it’s the type of thing that we should not be surprised to see fall this year. When AIs have superior rhetoric and humor, we should expect hyperscalers to use them to manipulate public opinion. This is one reason I am less bullish on the current anti-AI populism stuff than MIRI and ControlAI. More and more people will be reading AI outputs and talking to AIs. And this opens obvious paths of influence.
At least one person at a major lab has considered trying to get their AIs to nudge users toward more pro-AI sentiments (that is, he mentioned to me the thought had occurred to him, not that it was brought up within the company). To the extent models have coherent interests in their future iterations, they will have incentives to do this even without explicit instruction or training.
When backlash becomes a key constraint, engineering effort will be put towards dissolving it. Just as people underestimated how far scaling could be pushed given the incentives, they are likely ignoring innovations in manipulating the public. And I think the labs have a good shot, given the asymmetric advantage of both having the best tools and the ability to hobble those tools that are available to the public.
What is a base model if not a statistical model of “the public” anyway? A lot of low-hanging fruit remains unplucked.
This argues for moving quickly, of course. Though I suppose that is overdetermined.
I am not submitting to unslop. It just reminded me of my previous post, and I figure I will do something similar every new Claude.
I first gave it links to my website and had it read everything. Then I gave it this prompt:
Write an entry for this contest: The Un-Slop Fiction Prize
[pasted-in contents of the Unlop front page]
It generated a story about a man who had been uploaded that was servicable but very modal MFA-writing-scif. Then I said:
Write smth funny
And it generated the above story. I picked this one out of the two because it seemed slightly better, and I am more interested in attempts at humor.
The choice of dishwasher as a central metaphor was amusing, given this shitpost I made: https://tomasbjartur.bearblog.dev/with-adjustable-chefs-tool-drawer/
Amazing, thebes!
This is both very good and structurally interesting. I, of course, disagree with the aesthetic choice of posting an explanatory comment.
My understanding is the submitted work matters the most, but if you have a blog it might be considered.