niplav is @niplav.site · 24/09/2026Ask frontier LLMs for: • Solutions to shutdown problem • {Embedded, Cartesian} diamond maximizer • Solution to the Löbstacle • Open source game theory program equilibrium solutions 010
niplav is @niplav.site · 24/09/2026Is it common knowledge that Roman Yampolskiy is an AI Safety crank 100
niplav is @niplav.site · 24/09/20266fd7fbb3b81dbc0731ef61ae568db0c2383f844779c7ae556450e3f6393122a1 000
niplav is @niplav.site · 24/09/2026Convergent instrumental goals as another Darwinism-sized shift? 000
niplav is @niplav.site · 24/09/2026Concept: Moral internalism is true, but only if you have moral *knowledge*, if you have moral justified true belief (which humans have) one is not compelled 100
niplav is @niplav.site · 17/09/2026Recently: How much do I remember about the time when I formed this belief, so that I can treat events as being "priced in" already, vs. actually new? My memory isn't that good, so I don't have the full trace. Internet helps a bit 001
niplav is @niplav.site · 17/09/2026I do like that more people are told how the social network I belong to holds together, I like less that lots of to me irrelevant-looking details are emphasized. (Also the implication that the beliefs are wrong because they come out of a social network) 120
niplav is @niplav.site · 16/09/2026A distrust license is a situation where A has to give some level of trust to B, and the level of trust given is *also* a status signal, but B gives the license by pointing out the level of trust B thinks A should have on them given the evidence A has received about B so far. 000
niplav is @niplav.site · 16/09/2026Hm. AI speeds up history by a bunch, humanity would've died out in history far away, which is brought closer by said speedup. & species often get outcompeted by being reconnected with near-speciated rivals, I tried to get a baserate for this (endogenous vs. exogenous) but failed 140
niplav is @niplav.site · 16/09/2026It feels like ~60% of the Mythos transcript¹ is spent doing one of three things solving captchas getting a telephone number getting an email address If I were an AI agent these would be the three that'd most annoy me about humanity ¹: t.co/hktEfnHTR8t.cohttps://cdn.sanity.io/files/4zrzovbb/website/8359003bfb12a2f01ce84ad3df1d3a3e2f15a8eb.pdf 1131
niplav is @niplav.site · 14/09/2026Alright everyone time to start talking about UMAP plots of genomes and kin-selection mechanisms to even it all out 010
niplav is @niplav.site · 14/09/2026I miss agent foundations, it was fun before we were forced to do policy and evals (Opus 5 on LittleDarkAgeBench) 030
niplav is @niplav.site · 14/09/2026In Landworld, orthogonality is "false" because competitors with convergent instrumental goals always outcompete. Land loves them because they'd be gloriously efficient&beautiful in their function. But *also* my guess is that that world contains a lot of specialization/speciation 110
niplav is @niplav.site · 14/09/2026It seems like "we all get killed by our own invention" is less virtuous than "we all get killed by our enemy's invention"? 040
niplav is @niplav.site · 14/09/2026Big fan of the weird body {tilt, rotation, heaviness, expansion, valence-neutral contraction} feelings that arise in Body & Mind. "Yes, I'd like to feel like my torso is tilted/rotated by 25° towards my left, thank you very much" … "Oh, and weighing 300kg is also in the cards!" 000
niplav is @niplav.site · 09/09/2026Short story concept: Zoo with hyperecologies, i.e. terminal self-replicators. Signs with energy throughput, valence, complex systems classifications Something of a Strogatz+Egan+Land+Drexler+England story May also contain a wing with hedonium-variants See also: Dr. Diagoras (Stanisław Lem, 1961) 110
niplav is @niplav.site · 09/09/2026From asking Claude¹, looks possible to hard-brick GPUs with firmware access (p≈60%–70%) or even only software access (p≈15–25%). Accelerated aging (from years to months) through software has precedents, apparently! No nanotech required ¹: Query got me demoted to Opus 4.8forums.guru3d.com 110
niplav is @niplav.site · 09/09/2026Typology of multi-domain universes (From Grand Futures (Anders Sandberg, 2025)) 250
niplav is @niplav.site · 31/08/2026Note: A Toy Structural Causal Model For Psychological Agency niplav.site/notes.html#A... (Most important variable that could still be added: risk aversion?) 010
niplav is @niplav.site · 29/08/2026Okay, talking with Opus 5 about animal welfare makes it incredibly clear that the newer models have been getting less Good. Possibly a partial sysprompt issue on my side but damn this conversation made me actively dislike that model. 3150
niplav is @niplav.site · 25/06/2026Dzogchen has gömböced me… Sheesh, what a ceremony; a doozy of alcheringa guitars. My bonsai rigpa clanks in its satori quiver. Mack not with heroin— shiver, flirt at the obelisk up on this Jesuit Alhambra, up here with our geeky kamuy. Teotl whirs— wafts Mu 130
niplav is @niplav.site · 25/06/2026Over-agentificarion of nature≈animism, under-agentification of minds≈coercion 041
niplav is @niplav.site · 25/06/2026Short note in which I implement a set of commutative hyperoperators suggested by Ismael Ghalimi in an array programming language: niplav.site/hyperoperato... 000
niplav is @niplav.site · 19/06/2026Claude code can act as a cybersecurity specialist to check ones machine for exploits. It's even `cron`able. 281
niplav is @niplav.site · 19/06/2026Cults using LLMs for personalized recruitment? Doctrine that adapts with the mark? 060
niplav is @niplav.site · 12/06/2026I think it's almost entirely safe to give out base model or even non-RLVR model access by now. Misuse is very low/unlikely, the difficulty of eliciting anything useful is prohibitive. The only reason not to is worries about distealing, but even there that's less harmful. 000
niplav is @niplav.site · 12/06/2026Props to OpenAI for fixing GPTs personality. 5.5 is fine. I wonder if they simply exhausted the space of bad personalities. 000
niplav is @niplav.site · 11/06/2026As per Opus 4.8's analysis of my corpus, I use the word "plausibly" ~157× more often than standard English text 390
niplav is @niplav.site · 11/06/2026I would like to know how modern Online Reputation Management/skilled Public Relations firms operate. Claude doesn't know a good book on the topic, but this seems like an important case of adversarial epistemology to know about 120
niplav is @niplav.site · 11/06/2026Hard science fiction story concept, written in a series of fictional interviews, reports by the author, snippets from internet text: One day humanity receives a very loud and clear random-looking binary signal from some far-away part of space. Three years of false starts of interpreting the signal: 130
niplav is @niplav.site · 11/06/2026In modern media, power is often displayed as cool, collected, cunning, relaxed. In real life power may look bumbling, discoordinated, always almost failing? E.g. Elon Musk messes up a lot. As do Putin, Trump. Visible power is extremely out of equilibrium 040
niplav is @niplav.site · 11/06/2026HATETRIS for Grice-admissible ambiguities in the Twenty Questions Game/Who Am I? 121
niplav is @niplav.site · 01/06/2026Load-bearing "see full thread". Submissions welcome. @norvid-studies.bsky.social 270
niplav is @niplav.site · 01/06/2026If {Heidegger, Nagarjuna, Deleuze, Whitehead} are correct, then humanity was ontologically clueless before {Heidegger, Nagarjuna, Deleuze, Whitehead} were born 2101
niplav is @niplav.site · 01/06/2026Who decided to call it Internal Family Systems and not the Egosystem 081
niplav is @niplav.site · 01/06/2026Always fun to get obsessed with something and then discover a high-quality comment about it from 8 years ago, just to realize it's ones own 3410
niplav is @niplav.site · 01/06/2026dialectics.info scrapbook thread (I thought it was lost, but it's apparently still online!!!) 100
