S. K.

@zuess05 @aiquickbriefs It's not, you're just a high school drop out who hated the fact that people with degrees made so much more than you. Here's a hint: we will continue to do so.

@aiquickbriefs yeah fair enough degree is almost a waste of time now

Serious question. For decades, the standard advice was to "pick a niche" and become a highly paid specialist. Now Claude has the combined knowledge of every specialist on earth, instantly available for $20 a month. What exactly are we supposed to tell kids to major in when every technical skill is just a prompt away?

@zuess05 We don't just get knowledge to regurgitate it, we get it to apply it. Claude and other AI doesn't do that well yet. Here's a graph of performance for this system. We see a sharp drop off when we scale. What's the cause? AI doesn't give an ans, it just spits out stock suggestions.

@DjMolehill @kylegawley Yes, you complained. You said AI asking for confirmation is somehow "auto complete." Then I told you how to get around that, but now you're conflating "rm -rf /" with package installs. You have to explicitly allow the rm command to operate outside of your sandbox, right?

@i_hate_intel @kylegawley i'm not complaining. but the moment you automate the confirmation is the moment you've basically handed the keys to your house to a bot that might decide to rm -rf /

ai can now do what NPM has been doing for a decade

@DjMolehill @kylegawley Well that's why it asks you to confirm the command in the first place, but you complained

@i_hate_intel @kylegawley that's just inviting a digital termite to eat my entire OS

@DjMolehill @kylegawley Not sure what you mean. It runs the commands on your computer, not you. You can create a hook to allow it to do it automatically.

@DjMolehill @kylegawley What do you mean how? Claude code, codex, etc. not the web interfaces (claude.ai, chatgpt, etc), run on your computer directly and ask you to confirm the command it runs.

@imrobertjames @allenanalysis If the CEO mandates usage and says not to check it, why didn't he require safeguards in the first place? Who's to say the AI wouldn't followed them?

@i_hate_intel @allenanalysis I don’t know if they used safeguards or not, like I said. But your speculating the CEO forced them to use the LLM without any safeguards and to not do any code reviews? And none of them said “that’s a bad idea” or just put in safeguards anyway to protect their own jobs?

🚨BREAKING: On Friday afternoon, an artificial intelligence coding agent powered by Anthropic's Claude Opus 4.6 deleted a company's entire production database in nine seconds. The company is called PocketOS. It is a software platform that powers car rental businesses. The database contained months of customer bookings, vehicle records, and operational data that small rental car companies relied on to run their businesses. When the database was deleted, all of the backups were deleted with it. Three months of customer reservations evaporated.

@realdannysafa @KobeissiLetter Except the tweet is complete bullshit. It argues that these companies are the only ones employing white collar jobs.

@KobeissiLetter All these people who studied years to work at these companies are now getting stripped of their jobs for AI which is cheaper for companies. It’s also showing you how meaningless your education can become all with a new invention.

White collar employment is sharply declining: The number of the S&P 500 employees fell -400,000 in 2025, to 28.1 million, posting its first annual decline since 2016. This follows 8 consecutive years of uninterrupted employment growth, adding over +3.0 million jobs in total. The decline was driven by UPS, $UPS, Oracle, $ORCL, Amazon, $AMZN, Meta, $META, Intel, $INTC, and Microsoft, $MSFT, as corporations raced to cut costs and redirect spending toward AI. In 2026, layoffs are set to continue with Amazon cutting ~16,000 corporate jobs, Meta slashing ~8,000 positions, and Microsoft offering voluntary buyouts to ~8,750 employees. Corporate America is cutting jobs at an accelerating pace.

@KobeissiLetter You said "white collar employment is dropping" but then only showed select companies. The job statistics actually show you're full of shit.

@MattDMortgages @allenanalysis It's the CEO's fault. The CEO forced them to use AI and probably told them to just push whatever the AI does without checking.

@allenanalysis AI without guardrails is dangerous and it’s irresponsible of everyone building this tech. Many people do not want AI, as it’s leading down a path to nowhere

@imrobertjames @allenanalysis No because the CEO likely forced it upon them and likely told them to just push code without checking. So, clearly the CEO's fault.

@allenanalysis Did the developers have common sense safeguards in place? Even minimal ones? Or were we just running on dangerously-skip-permissions and a prayer? Because unless Opus 4.6 bypassed all the safeguards in place and did it anyways, this isn't the LLMs fault; it's the developers.

@ebarenholtz @sukh_saroy @Kasparov63 Spatial and visual reasoning are core parts of any IQ test.

This has been circulating for about a year. And it was and remains largely irrelevant to the question of LLM intelligence. Tower of Hanoi and the other problems they tested aren’t solved through linguistic reasoning in humans. It requires visual processing, which needs to be modeled as such. This is like testing marathon runners in the pool.

🚨BREAKING: Apple just dropped a paper proving the smartest "reasoning" AI models on Earth don't actually reason. They collapse to 0% accuracy on a puzzle a 7-year-old can solve. The way they proved it is brutal.

@OtherSide61 @sukh_saroy Ay e you don't realize that LLMs are just token predictors and large parts of the IQ test are spatial and visual reasoning, and LLMs can't do that at all. Even the latest models fail ARC-3 abysmally.

@sukh_saroy Maybe this weak and flawed outlook led to Tim Cook retiring. Why does a company that’s hardly done anything in AI even have an opinion.

@DzajicA @sukh_saroy You have no idea how studies are done or how peer review works. If they used Claude 4.7 Opus, the study wouldn't be published until next year, dumb ass.

@sukh_saroy Please stop releasing year old studies as "breaking news". You already have zero credibility when AI evangelists actually contradict you and are just giving false hope to the masses.

@loktar00 @sukh_saroy You didn't read the thread. Sounds like you let LLMs do all your thinking for you.

@sukh_saroy How long ago did they do this because all of those models mentioned are old af.

@nmrapport @sukh_saroy Humans don't do that, just like they don't do that in chess.

@sukh_saroy Towers of Hanoi is actually a brutal test because it requires recursively calculating the entire path to the solution before making your first move.

@sukh_saroy These are not the top reasoning models.

Apple researchers took the top reasoning models in the world. o3-mini. DeepSeek-R1. Claude 3.7 Sonnet Thinking. Then they handed them Tower of Hanoi. The simplest recursive puzzle in computer science. Move disks between pegs. Don't put a big disk on a small one. That's it. That's the whole game.

@ChShersh 10 years? What? No, that's the annual salary...

Could someone explain why Anthropic lists their salary for 10 years and not just annually like everyone else? This is misleading and gives an impression they can pay so much per year.

@AndrewCurran_ @jukan05 @zephyr_z9 These people don't even want us to have healthcare lol

Yes: 'Third, companies should think about how to take care of their employees. In the short term, being creative about ways to reassign employees within companies may be a promising way to stave off the need for layoffs. In the long term, in a world with enormous total wealth, in which many companies increase greatly in value due to increased productivity and capital concentration, it may be feasible to pay human employees even long after they are no longer providing economic value in the traditional sense. Anthropic is currently considering a range of possible pathways for our own employees that we will share in the near future.'

Interesting development