Bornach

@bellegraylane@universeodon.com @cstross@wandering.shop Musk merged Xitter with xAI to justify its high valuation to investors as an AI company now. The same crap with Tesla being rebranded an AI robotaxi and humanoid robot company. So makes sense to pull the same trick with SpaceX to gullible investors. That it's really an AI company so that SpaceX can afford to bail out Tesla when it buys all those unsold Cybertrucks. Won't be surprised when Neuralink is touted as an AI company next

@keoni @camertron Once worked for a narcissistic bosshole who told me how the Internet was such a game changer, and implied his bullying of staff to work harder than before was somehow justified. For similar reasons sociopath bossholes love AI because it justifies pressuring their employees to adopt the "hardcore" work culture.

@climbertobby@chaos.social @amsomniac@mastodon.mit.edu @lynnesbian@fedi.lynnesbian.space Only missing the emoji 🐢🔪🏃‍♀️

@RunRichRun @mattsheffield @jwz Also derivative of Sun Yuan & Peng Yu who did the same concept better in 2007 with their "Old People's Home" https://youtu.be/XEoeHt07HAw

@GhostOnTheHalfShell @mattblaze Actually I'm proposing a drinking game for every time I see that "What's Going On With Shipping" host do an impersonation of the FIFA Peace Prize trophy https://masto.ai/@bornach/116262514467727863

@qgustavor@urusai.social @oatmeal@kolektiva.social @tante@tldr.nettime.org "free" alternatives With Sora gone, there is little reason for Google to keep their own AI video generator money pit running. Those unrealistically low tiers will soon vanish. The same applies to code generation. Even $200/month is still only being sustained by continuing investor funding rounds. https://youtu.be/oAbpVCn-Ox0

@oldgeek @lucien @cstross At 1:14:17 in the latest rant by @TechConnectifyhttps://youtu.be/KtQ9nt2ZeGM?t=1h14m17s Imagine if government never built power lines to rural areas. They'd probably be singing praises to orbiting space lasers beaming them energy at huge expense or the delivery robot drones dropping off daily fuel shipments for their generators.

@nblr @carstenfranke @mattblaze Nothing stopping FIFA awarding it posthumously

@simondassow@masto.ai @toni@zug.network @vkc@linuxmom.net Yup. Many devs have a libertarian free market competitive streak to them. They only saw acquiring coding skills as a means for competitive advantage. Their dopamine hits came not from creatively solving problems, but from seeing how they were getting ahead of the competition. That's why AI slop machines would be so appealing to them.

@maxthefox@spacey.space @kageRapaz@tech.lgbt @timnitGebru@dair-community.social Which is why even path finding algorithms to move a NPC in a game from point A to point B while avoiding stationary obstacles is considered "AI"

@lynnesbian@fedi.lynnesbian.space As Roy Batty spends his last few minutes pushing Deckard's car 50m to the car wash

@ngaylinn@tech.lgbt @DrewKadel@social.coop Actually we as a species have had to deal with that before. We call them grifters.

@emlove@kolektiva.social @timnitGebru@dair-community.social Only big computers in data centers count though. AI cannot run on your home PC. If the public wants AI they must allow AI tech billionaires to build a giant data centre encroaching on your back yard.

@graydon@canada.masto.host @mikemccaffrey@wandering.shop @adaliabooks@ohai.social George Pearkes over on Bluesky claims this purported explanation is nonsense https://bsky.app/profile/peark.es/post/3m6b6u73vw22h

@mina@berlin.social @adaliabooks@ohai.social @cstross@wandering.shop See this debunking of that "success of ai" framing https://bsky.app/profile/peark.es/post/3m6azdw5hwk2p

@ngaylinn@tech.lgbt LLMs are never your friend They are always in what can best be described as "the grifter" mode. The entire training regime of a generative AI chatbot is geared towards getting one thing, an upvote from a human rating the quality of the conversation Admittedly this is an over-simplification. Reinforcement Learning with Human Feedback involves training a reward policy - a 2nd neural network that is ultimately responsible for rewarding the chatbot for giving "good" responses.

@mergesort@macaw.social @rodhilton@mastodon.social @mmasnick@mastodon.social I now have a tab open on https://macaw.social/public/local I worry they might overdose on schadenfreude

@TaoBear@mastodon.social @cindyweinstein@mastodon.world They were victims of this hack? https://youtu.be/9FdHq3WfJgs

@Sector9 @NEPTOON @zoeschiffer The only problem is we're watching all this happen from the vantage point of one of the lifeboats which is already at capacity

@mdreid@mastodon.social @timnitGebru@dair-community.social @emilymbender@dair-community.social @alex@dair-community.social @baldur@toot.cafe Trouble is marketers and AI techbros are deliberately muddying the distinction between gen- and nongen-AI. We now get ridiculous claims that the entire protein folding problem has been solved by a LLM (or by ChatGPT, or by OpenAI, or by Sam Altman) because a LLM can be implemented using a transformer and DeepMind AlphaFold had also used a transformer in just one of among the several ML techniques it uses. https://medium.com/meta-multiomics/alphafold-is-not-an-llm-model-0646e7c3c01c

@tristen@illo.social @vkc@linuxmom.net Same AI techbro probably thinks no one likes making music either https://www.musicradar.com/music-industry/its-not-really-enjoyable-to-make-music-now-says-the-ceo-of-an-ai-music-making-platform-and-he-wants-to-help

@GordanKnott@mastodon.social @ngaylinn@tech.lgbt @DrewKadel@social.coop Yet another flawed benchmark in which the LLM very likely memorised the answers https://wandering.shop/@janellecshane/110104164829618120 Without any knowledge of how much the training dataset was contaminated by the medical exam questions/answers (and OpenAI's own whitepaper admits there is contamination) https://youtu.be/PEjl7-7lZLA?t=4m0s we cannot really know how it would perform in the real world if say a novel virus were to start spreading

@maxpool@mathstodon.xyz @unchartedworlds@scicomm.xyz @brucelawson@social.vivaldi.net Many were skeptical of Tesla's $56B pay package in 2018 https://www.cnbc.com/2018/03/21/tesla-shareholders-approve-elon-musks-multibilion-dollar-compensation-plan.html To hit those targets he just has to find more gullible tech bro investors and keep inflating the hype bubble with ridiculous unfulfillable promises. Failing that just sell any unsold Cybertrucks and Optimus robots to xAI and SpaceX at massively inflated prices. https://electrek.co/2025/10/13/elon-musks-spacex-and-xai-are-buying-teslas-unsold-cybertrucks/

@maxpool@mathstodon.xyz @unchartedworlds@scicomm.xyz @brucelawson@social.vivaldi.net Revenue is irrelevant when you have the government in your pocket https://uk.news.yahoo.com/pentagon-reportedly-award-spacex-2-210019639.html And xAI is backed by Nvidia through a magic money tree called a Special Purpose Vehicle https://www.datacenterdynamics.com/en/news/nvidia-to-invest-in-xai-special-purpose-vehicle-cash-to-be-used-to-buy-nvidia-gpus/ "Jensen Huang said that his only regret with xAI and Musk is that he “didn’t give him more money.” "“Almost everything that Elon’s part of, you really want to be part of as well,”"

@AnguaDelphine@mk.absturztau.be @kyonshi@dice.camp @stopthatgirl7@famichiki.jp I guess it depends on the civilization from which one is viewing the said progress. Those being exploited, bullied, suppressed, and subjugated by the Roman empire might have viewed seeing it collapse as progress, for example

@natcbr@mountains.social @tante@tldr.nettime.org I guess OpenAI found it prohibitively expensive to fine-tune/retrain/police Sora in order to satisfy Disney's legal restrictions: no sex, no drugs, no famous actors voicing their characters

@modulux@node.isonomia.net @timnitGebru@dair-community.social Even more reason to avoid using the term AI without spelling out what specific ML technique one if referring to. People disagree over what falls under the broad umbrella of AI.

@jfparis@rouge.eu.org @levelbot@mastodon.socialhttps://makeagif.com/gif/airplane-unbelievable-bullshit-tpy8GR

@maxthefox@spacey.space @meltedcheese@c.im @timnitGebru@dair-community.social In the last year? There was a lot of research into combining LLM + tool use. One of the tools might have been a just-in-time compiler of a script regurgitated by the LLM in order to answer a prompt. Running the compiler behind the curtain generates an accurate numerical answer in place of the inaccurate one a LLM operating alone would likely have concocted.