AntiDr @antidr.bsky.social · 09/10/2026And like, even if you only had depth... the cat would be a slight anomaly relatively to the points around it which would be relatively easy to pick out too 030
AntiDr @antidr.bsky.social · 09/10/2026He's also seemingly ignoring that you don't even need to run any segmentation on the point cloud to detect objects? The cat would definitely show as a separate object from the ground even with some relatively simple clustering. 140
Reposted by AntiDrAdam Cook @motorcityadam.com · 08/10/2026Lol. I bet Yun-Ta thinks so. Anyway. Pretty much demonstrates that the remaining personnel on Tesla's Autopilot Team went to the Elon Musk School of Robotics - which only contains incompetence and lies. If we actually had auto safety regulators anywhere globally, hard questions might be asked. 5628
Reposted by AntiDrEmily M. Bender @emilymbender.bsky.social · 07/10/2026I saw someone breathlessly talking about this as "AI produced proofs that will take humanity years to understand" ... which just told me that they really don't know what scholarship is, in math or otherwise. 1152
AntiDr @antidr.bsky.social · 07/10/2026"Their version of a primary", not a primary just the election itself 😉 010
AntiDr @antidr.bsky.social · 06/10/2026Complete non-answers, and must say the Guardian isn't really providing any analysis. Wonder why? Oh, nothing to do with that partnership with OpenAI I assumetheguardian.comOpenAI delivers a mea culpa to the Australian government in person – but answers still eludeJason Kwon’s answers were provided in a helpful, quiet and calm manner. For a company under intense scrutiny, there were no missteps 000
AntiDr @antidr.bsky.social · 06/10/2026One cannot overstate how fucking bad SpaceX IVA suits are lmao... 100
AntiDr @antidr.bsky.social · 06/10/2026Let me correct that title for you: "French police attacks protesters as per its usual violent ways" 000
AntiDr @antidr.bsky.social · 06/10/2026I've seen him have a few bad takes before, I'm pretty sure we had a a fight once here about whether there is any usefulness to genAI lol 000
AntiDr @antidr.bsky.social · 06/10/2026That doesn't make them *all* computational linguistic, or, especially, ethical 120
Reposted by AntiDrWolfie Christl @wchr.bsky.social · 02/10/2026LLM chatbot surveillance advertising is going to be next-level disastrous, on every level. 02013
Reposted by AntiDrWolfie Christl @wchr.bsky.social · 02/10/2026Plus, OpenAI's new integration with LiveRamp (acquired by Publicis) enables profiling and targeting using population-scale identity records linking names, postal, email, phone, and digital IDs associated with devices, browsers and platform user accounts: liveramp.com/blog/bringin...liveramp.comBringing the Power of RampID Globally to Custom Audiences in ChatGPT AdsBringing the power of RampID globally to Custom Audiences in ChatGPT Ads 1159
Reposted by AntiDrWolfie Christl @wchr.bsky.social · 02/10/2026The web and mobile apps of major conversational AI engines transmit prompts, titles and conversation permalinks to advertising data firms, alongside device IDs, hashed email and other persistent identifiers. Comprehensive analysis, like always from this team: jorgegarciaherrero.com/wp-content/i... 26767
AntiDr @antidr.bsky.social · 06/10/2026We really need to kick neolibs out of the RBA so it can try to adress actual causes of inflations instead of staying stuck in its brainrot 010
AntiDr @antidr.bsky.social · 06/10/2026Also funny to think that is a new thing when the US EV market has been lagging behind the other big markets for quite a few years now 000
Reposted by AntiDre.w. niedermeyer @niedermeyer.online · 05/10/2026I don't know who needs to hear this, but the autonomous vehicle developers who most vocally argued that society must accept a certain number of deaths to enjoy the technology's safety benefits (Tesla & Uber ATG) were the only ones who killed people and never actually delivered on their promises 161596367
AntiDr @antidr.bsky.social · 06/10/2026It's always "funny" when other countries are following failed US ideas years later, seemingly not realising how they failed. Europe is usually quite "good" at that 010
AntiDr @antidr.bsky.social · 05/10/2026Russians were also purchasing the vast majority of the output of sweet ("doux") champagne from France while everyone else only want brut or demi-sec 010
AntiDr @antidr.bsky.social · 05/10/2026It's pretty clear to anyone but deep stans that SS only exists to launch Starlink, and the fact that SPX seems to actually be real about pretty much withdrawing from the normal launch market by retiring Falcon supports that 010
AntiDr @antidr.bsky.social · 05/10/2026Well I'd rather have them die of being stupid than because of a false sense of how capable the system actually is. Of course no fatalities is the best option 000
Reposted by AntiDrEric Lagadec @ericlagadec.bsky.social · 02/10/2026Survol de Jupiter par la sonde Juno, révélant d'immenses orages, plus grands que la Terre ! 1023648
AntiDr @antidr.bsky.social · 03/10/2026> reliability Sloppiness has been increasing with F9 recently, crew dragon had many similar issues to Starliner (most of them much less publicised for "some reason"), and the idea that Raptor/Starship is reliable is very, very funny indeed 100
AntiDr @antidr.bsky.social · 03/10/2026> performance Their rockets are nothing special compared to others? Are you one that believe the fake FH performance of 64t is real and not a theoretical number rendered moot by the structural limits of FH being the same as that of F9? 100
AntiDr @antidr.bsky.social · 03/10/2026> price They are not the cheapest and for NASA the costs of launches is actually trending up? 100
AntiDr @antidr.bsky.social · 03/10/2026Aside from Shotwell being a lunatic liar, the cost of Starliner and Crew Dragon are pretty much the same (6 ish billion each) once you account for the fact that cargo dragon need to be counted under the crew variant cost (since Starliner didn't have the benefit of an existing platform to work from) 200
AntiDr @antidr.bsky.social · 03/10/2026> I wouldn't trust Boeing to run the astronauts to the pad. Hey boofhead, who do you think is the main contractor for SLS? 000
AntiDr @antidr.bsky.social · 03/10/2026I mean a lot of knives out is also not justified and just fans or propagandists for *a certain company* spreading a specific agenda there because it's not really proportionate to the actual issues overall 100
AntiDr @antidr.bsky.social · 03/10/2026Sex in robotaxi is whatever. Sex at the wheel of a level 2 system (AP/FSD) on the other hand is a real safety issue 👀 110
Reposted by AntiDrEd Zitron @edzitron.com · 02/10/2026More than a trillion dollars and four years into the AI bubble, it's not unreasonable to expect that we'd have solid data showing that AI was this major economic agent of change. www.wheresyoured.at/premium-how-... 110110
Reposted by AntiDrEd Zitron @edzitron.com · 02/10/2026Premium Newsletter: How Has AI Changed The Economy? Not very much outside of GPU sales and data center construction, with AI software sales and GPU rentals having a negligible effect, even as BLS data overemphasizes software's contribution to real GDP. www.wheresyoured.at/premium-how-...wheresyoured.athttps://www.wheresyoured.at/premium-how-has-ai-changed-the-economy/This week, the Wall Street Journal ran an alarming illustration of AI’s potential share of GDP that I’d argue did more to muddy the waters than actually telling anyone anything, mostly because its… 631750
AntiDr @antidr.bsky.social · 02/10/2026High level clown shit in the centre-left primary debate. Faure does come out as the least bad candidate here, Glucksmann as Macron Lite 000
AntiDr @antidr.bsky.social · 01/10/2026Guess that could happen indeed, but it will probably still have more usage friction than just loading chatgpt now lol. Unrelated to image generation but using a bot as a search engine is one of the "normie use-case" that probably is the most vulnerable to OpenAI/Antro dying 110
AntiDr @antidr.bsky.social · 01/10/2026I don't even think coding assistants are safe because even if people think they are useful (freaks...) they are still mostly coming from the same hyperscallers than won't survive the bubble popping! 010
AntiDr @antidr.bsky.social · 01/10/2026"neural nets" in general if we mean ML are not going to disappear of course and probably also not transformers since they are used in other things than LLM. Specifically what can be "undone" is LLM-based public facing systems, because in most cases they require a massive infra to run 110
AntiDr @antidr.bsky.social · 01/10/2026Yes but the difference in my opinion is that locally run models are one step too far for most "normies" so once chatgpt/claude/etc are not accessible anymore you are going to see, for example, the end of the epidemic of ai generated small business posters & such 200
AntiDr @antidr.bsky.social · 01/10/2026So IMO functionally for normies it's actually possible to "unring the bell" to a significant degree 010
AntiDr @antidr.bsky.social · 01/10/2026Here to stay doesn't mean usage on mass scale in the general public will stay though. Once the main apps are dead because OA and Anthro ran out of money you start to get too much friction for most people 210
AntiDr @antidr.bsky.social · 01/10/2026"à ses yeux ça sera toujours les plus gentils/meilleurs" jusqu'au jour où quelqu'un d'autre que Melon (par example Jeff) le paiera plus, bien sûr 000
AntiDr @antidr.bsky.social · 01/10/2026Anyone paying attention to the steady degradation of SpaceX in the past few years knows "exceptional ability" is def not present there 020