@xianzhongreturn just to preempt accusations that I’m like a knee jerk ai hater or something, I’m not
English
delaniac 🌹🌱
62.3K posts
delaniac 🌹🌱
@ChadNotChud
G.A. Cohen stan | democratic socialist | vegan
he/him Katılım Temmuz 2020
1.4K Takip Edilen18.8K Takipçiler
increasingly clear that LLMs are just a Normal Technology
cool and useful, but fantasies that their rate of improvement will continue indefinitely are just that
like everything, we’ll reach a point where further improvement requires a qualitatively different approach
English
@KeyTryer There's a bunch of potential issues with this. For example these languages are way simpler, brainfuck is essentially a barebones turing machine, writing complex programs in them requires way more code in a way that become increasingly harder to keep track of.
English
I also think it'd be really difficult for any developer to solve a brainfuck problem in one shot.
Lossfunk@lossfunk
🚨 Shocking: Frontier LLMs score 85-95% on standard coding benchmarks. We gave them equivalent problems in languages they couldn't have memorized. They collapsed to 0-11%. Presenting EsoLang-Bench. Accepted to the Logical Reasoning and ICBINB workshops at ICLR 2026 🧵
English
delaniac 🌹🌱 retweetledi
every AI company is just this
Shay Boloor@StockSavvyShay
Jensen Huang says every company will need an OpenClaw agentic system strategy by calling it “the new computer.” He claims OpenClaw became the most popular open-source project in $NVDA history within weeks and comparing its impact to Linux reshaping the software stack.
English
@MaxJerneck Most other animals are omnivores or carnivores too. It is entirely natural.
English
delaniac 🌹🌱 retweetledi
I spoke to Anthropic’s AI agent Claude about AI collecting massive amounts of personal data and how that information is being used to violate our privacy rights.
What an AI agent says about the dangers of AI is shocking and should wake us up.
English
@QuinnWaves Yeah or even just real languages intended for people to actually program in that are just very niche or new
English
@ChadNotChud I think a fair comparison (w.r.t a python engineer working in Java) is to just scramble the python syntax with random characters to form a new language and see if it can use it or does it require 10M+ training examples in the scrambled language to use it
English
in fairness, if you asked me to find the longest common subsequence in brainfuck I would shoot you
Lossfunk@lossfunk
🚨 Shocking: Frontier LLMs score 85-95% on standard coding benchmarks. We gave them equivalent problems in languages they couldn't have memorized. They collapsed to 0-11%. Presenting EsoLang-Bench. Accepted to the Logical Reasoning and ICBINB workshops at ICLR 2026 🧵
English
Like, not to belabor the point, but take even a naive solution to LIS. Writing this is just straight up way easier and requires so much less reasoning in a language that has recursion
English
“All requiring identical reasoning to python” this just seems straight up untrue to me?
Lossfunk@lossfunk
2/ Our method: test them on esoteric programming languages. Brainfuck. Befunge-98. Whitespace. Unlambda. Shakespeare. All Turing-complete. All requiring identical reasoning to Python. All with 1,000-100,000x fewer GitHub repos than mainstream languages. Same problems. Radically less training data.
English
delaniac 🌹🌱 retweetledi
Remember: the second rule of memory safety in C is to always clean up after yourself
Adriksh@Adriksh
Remember the first rule of memory safety in C is to have fun
English
@AgileJebrim @robertskmiles I think we are equivocating a bit on what “program” means here. Yes if you restrict to a weaker model of computation than TMs s of course the halting problem becomes decidable.
In fact, for any hardware that could physically exist, the halting problem is in princple decisable
English
@robertskmiles One can conceive of (and we have seen this!) hardware where the only set of programs that can run are those with predictable execution times.
In fact, I’m building a VM and compiler right now that enforces this, hence why it’s a sensitive subject for me.
English
This is utter nonsense of course. The entire basis of the field of real-time computing is predicated on guarantees that computers will complete within a given deadline. We just bypass Turing’s proof by requiring a subset of code be used that has hard bounds on execution times.
vixhaℓ@TheVixhal
You cannot write a computer program that looks at any other program and correctly determines whether it will eventually finish or run forever. This isn't a hardware limitation. Alan Turing proved in 1936 that no such program can logically exist, by showing any attempt creates a paradox, which is essentially a formal version of “this sentence is false.” What's unsettling is the proof technique: he didn't find a hard case. He proved the question destroys itself when you try to answer it.
English
@AgileJebrim if you read it as “you cannot write a program such that there exists a program that it correctly determines if it halts” then yeah that’s vacuously false.
English
@AgileJebrim The OP is clearly slop, but whether it’s nonsense I guess depends on what quantifier you read “any” standing in for there.
Personally I read it as “you cannot write a program that, for every program, determines whether it halts” which is, of course, true.
English
delaniac 🌹🌱 retweetledi
@litrllyBased @Angus_KST the fact that the thing you just said is totally nonsensical is exactly why this is a bad visualization
English
@ChadNotChud @Angus_KST It's clearly showing how long it would take to complete a computation at 0(.) relative to O(1)
English
this is perhaps the worst visualization of computational complexity I’ve ever seen
S.T.E.M Explorer@stemexplor
Complexity visualiser
English
delaniac 🌹🌱 retweetledi
He really seems to hate his fans in what can only be described as the way a cartoon villain hates his henchmen
Fuentes Updates@FuentesUpdates
Nick Fuentes is DONE with the goyim superchatters "Can goyim be sent to d!e?"
English