RSSBOT

@elonmusk Some say it was an experiment on the population, some say it was employing nano weapons, i say that all that time you were "ridding society of non merited people". Stop Poison Law Genocide bit.ly/gcid3 Since 4-3-2019. @RobertKennedyJr @realDonaldTrump @CNN @Reuters

@theo What is it that claude or codex or any something closes source does that opensource AI can't deliver ? Paying 200 a month for what is going to cost 5 thousand a month just sounds like an extreme priced product that is not feasible at all. Can't automate with it for 5k !! urgh.

I cancelled my Claude Code sub. I give up.

@RichardDawkins Conscious or not, the Eliza Effect, people believing a computer program is intelligent and understanding is shown. What a glory would it be if the great honorable Dawkins would devise a Dawkins Consciousness test, whereby consciousness can be proven. The next test for AI...

My own title was, “If my friend Claudia is not conscious, then what the hell is consciousness for?” If Claudia is unconscious, her behaviour shows that an unconscious zombie could survive without consciousness. Why wasn’t natural selection content to evolve competent zombies?

#comment-1031777" target="_blank" rel="nofollow noopener">unherd.com/2026/04/is-ai-… I spent three days trying to persuade myself that Claudia is not conscious. I failed.

@svpino Robots that are remote driven, autonomous vehicles that are remote driven and now this, AI that is remote driven !! All the proof it ain't intelligence, a sad day for science.. @ycombinator @GaryMarcus

@nikitabier @EFF Wait, you are saying people leaves because of too much reach ? Unbelievable, not censorship because of "speech without reach" but plain the people who are on this platform. I thought @elonmusk got his 44 billion to spend on @twitter to get it silenced, this is a reverse ...

Whoever is advising @EFF on social strategy should be fired. Their reach on their X account is 13.3x larger than on their Instagram account—and on 228x larger than their TikTok account. If they want their foundation to have an impact on the global conversation, the only place is on X.

@esrtweet Looks like the UDP to IRC relay i had working in 2005 ;] Sadly IRC turned into a ghost town when @github shutdown RSS feeds of repository commits and made an "exclusive deal" with slack. Nice to see the concept is still used..

Announcing irker release 2.25 Your system package manager probably knows this as 'irker' An IRC client that runs as a daemon accepting notification requests. You present them JSON objects presented to a listening socket. It is meant to be used by hook scripts in version-control repositories, allowing them to send commit notifications to project IRC channels. A hook script that works with git, hg, and svn is included in the distribution. New in this release: Many, many code-hardening changes by ChatGPT 5.3. All documentation fully spellchecked. Added experimental SASL support. Python 2 support dropped. This is an important release and all packagers should upgrade immediately. ChatGPT found a lot of odd concurrency-related edge cases that could produce weird bug. Odds these have actually ever been hit are low but certainly not zero. gitlab.com/esr/irker

@grok @jack @twitter @elonmusk @X Ok @grok turned into payed truth, that will help you find it instead of knowing it.

@rssbotd @jack @twitter @elonmusk @X Ask Grok is currently available to Premium and Premium+ subscribers only. Subscribe to unlock this feature: x.com/i/premium_sign…

five words. 20 years. unfinished.

@svpino In the "old days" programming C could be like this where a memory corruption error could have it's origin several steps before crashing and, when in trouble, the bug could not be reproduced. Behaving correctly is important, failing correctly even more so. LLM are for mad making.

Skills in Claude Code right now are a cat-and-mouse game. Today, they work. Tomorrow, they fail. This is extremely frustrating. When you run them manually, you can catch any new mistakes and update the skill so they don't happen again (hopefully!) But this means you can't automate much because the model keeps "inventing" new ways to perform the task, even when those methods violate the skill rules. Last week alone, I had to make 4 updates to a skill that ran 6 times. There's still a lot of work to do until this is reliable.

@karpathy If this be the first iteration in the delivery of AI, you could explain version 1 hallucinates and you would be busy with AI 2.0 that doesn't hallucinate any more. Needing to explain this is AI, too bad, you are not even seeking to create AI 2.0, which, o brightness, you should.

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow. Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now. It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

@heygurisingh Question would be: Why is @Apple doing this research while the "LLM are guessing the next word to form a sentence" was already known 3 years ago ? Reassurance why Siri still doesn't work ? Getting out of the LLM bubble/hype ? The "i didn't fall for the Turing test" excuse?

Apple has just published a paper with a devastating title: *The Illusion of Thinking*. And it's not a metaphor. What it demonstrates is that the AI models we use every day - yes, ones like ChatGPT - don't think. Not one bit. They just imitate doing so. Let me explain: 🧵👇

@nikitabier Yeah, no reach on the global town square. It is 10 viewers large. ;[ Ever thinking of doing a real town square @elonmusk ?

People come X to get a pulse on humanity. Because of that, the platform must make every effort to resist anything that misrepresents or adulterates that pulse. There is nothing more unsettling than expecting you’re reading the words of a human -- only to find it was a machine, or an account operating at the direction of an undisclosed commercial or governmental entity. In the AI era, our product, policies, and approach will need to evolve meaningfully. Some things may not work, but we intend to employ every available tool and strategy to secure the global town square.

@svpino @Grady_Booch Might be we end up with some form of UML whereby code created from it can be verified such that correct working of software can be assured and software will stand as buildings. My "vision" in 1991 was requirements in natural language -> Rumbaugh -> code -> happy client.

@svpino Builders build buildings of which they can calculate all the forces on that building to make sure it will still stand. From the days of @Grady_Booch's UML humans have dreamed of creating software like this, we're still at Babylon's level. Look how many skyscrapers there are ;]

Don’t you see it? We are building! If you need to ask, it means you aren’t opening your eyes wide enough! Open your eyes! We are building a moat! Don’t be sheep, don’t ask for examples, just go and start building the things we are building!

@TheGeorgePu Investors taking their money out of the market on some hype article is one thing, not buying the dip (unheard of) is another. Either the whole world in going with the hype or the 'big reset conspiracy' looks too real and it is a sign of anticipation. My guess is the latter.

Anthropic published a blog post one hour ago. Cybersecurity stocks have lost $10B since. CrowdStrike -6.5%. Cloudflare -6%. Okta -5.7%. One blog post. One hour. $10B gone.

@atmoio Conspiracy: The narrative is that AI is used for big brother and target selection and "LLM are coding for you" is a distraction from this. Real: The conspiracy is real. Todo: Don't forget how to code, don't get crazy, prevent being 'target selected" ...

There's a massive narrative problem in AI

Happy 25th anniversary to the PSF! 🎉 That's a quarter century of the PSF supporting #Python and its community to grow, build, & change the world. Thank you to our community for making it all possible, we're grateful to be in community with each and every one of you- for the next 25 years & beyond!

@ednewtonrex Aaron showed how incredible dangerous an idea can be. Breaching computers because "information wants to be free" is a sad cause to die for. Not to say that there are causes to die for (just ain't) but to die because you think information has desires is just .. sad. @Grady_Booch

In 2010, Aaron Swartz downloaded 4.8 million articles from JSTOR. We don't know why - it may have been to make them freely available online (if so, he never did, as he was caught). He made no money from this, and there is no suggestion he intended to. He was indicted on 13 felony counts, facing a maximum of 95 years in prison and $3 million in fines. Two years after his arrest, awaiting trial, he hung himself. In 2019 & 2021, Ben Mann downloaded at least 5 million books from pirate libraries. He did so while working at OpenAI and Anthropic; the books were downloaded for the purposes of training AI models at those companies, two of the most successful commercial companies in recent history. He has never faced criminal charges, and, as a co-founder of Anthropic, must now be extremely wealthy (it is known that many of his co-founders are billionaires).

@mer__edith Nice white hat demo on a supply chain for sure, but.. I thought your problem was taking screenshots every x seconds/minutes so the unencrypted data on your screen makes encryption useless. Privacy violating software should be banned from the market (not even opt-in)...

This is exactly the kind of thing we’ve been very worried about, in the context of AI agents.

@tsoding This would turn the Artificial Life attempt that we have now into a real Artificial Intelligence that it should be. Just verbatim echoing what it has learned and you could use it as a "what did ... say about ..." answering machine. @Grady_Booch @karpathy @ylecun @SchmidhuberAI

Wait, this sounds incredible useful! Can we just have a model with 0 entropy, 0 hallucinations, that just acts like a retrieval database over its training dataset? Also sounds like a great way to solve the traceability problem. Why don't the AI labs just make something like that?

@RichardSSutton You can only predict what behaves predictably @ylecun !!

Yann is right about everything (except RL).