DigitalCharityLab

we're reaching john stuart mill levels of absurd "thought experimetns are just as good as real experiments" scientism with this one

it’s pretty incredible how unusable search is on almost every part of the internet now. a product that worked almost without fail for 7-10 years is just utterly incoherent. it’s as if every search feature on every website simply does not understand the words you’ve typed

Executives are used to sycophantic behavior, their power over their employees means they’re constantly told their ideas are great and they think they’re the smartest person in the room Normal people find AI sycophancy off putting, executives think it’s normal and validating

TMZ is now chasing politicians & getting engagement they haven’t seen in years. Their politician posts get like 10,000x more engagement than their celebrity posts. I predict they’re going to pivot to politics, & social media censorship may have been the trigger. How funny.

if a man rapes a girl, he’s still allowed to compete but yes please focus on trans women again and again and again

Hate to think of how many students have suffered and job applications have been ignored because the software that does stuff like this basically doesn’t work.

She’s monetizing her dad’s transphobia to raise money for trans kids. Absolute queen

AI is inevitable. You can tell by how no one wants it and it’s losing money left and right.

made my computer dramatically play BBC news music before every meeting

The most successful thing capitalism ever did was make labor invisible. Just consume, it doesn’t matter how it got here. It doesn’t matter who made it or the communities that are suffering from just being near the factories. You deserve a little treat!

I wonder if there are any other tech products being breathlessly hyped as the all-consuming paradigm shift of the future that will look similarly cringe and dated in ten years

Is organic social completely dead? 😵 Plus - quizzes, AI readiness checker and more - mailchi.mp/digitalcharity…

A lot of the people currently shouting that generative AI is the future whether you like it or not said the exact same shit about the Metaverse

Like 10 years the American Christian right held a fucking conference where they discussed how hard they found it to break into the UK market and they solidified around anti-trajs bigotry as their way and committed a tonne of money. They didn't even hide it!!!

The web never should have progressed beyond this

It is genuinely wild that a company with a user base of *creatives*, thought it could wedge AI into its products without backlash. Yet another example of terrible judgement from tech about AI.

Solidarity to the comms team at @DublinZoo - even when a backlash is obviously egregious, astro-turfed nonsense, it’s still horrible to have to deal with it

OpenAI just published a paper proving that ChatGPT will always hallucinate. Not sometimes. Not "until the next version." Always. They proved it mathematically. And three other top AI labs confirmed it independently. Here's what the research actually shows: Even with perfect training data and unlimited compute, LLMs will still fabricate answers with complete confidence. This isn't a bug in the code. It's fundamental to how these systems are built. The numbers are wild: → OpenAI's o1 model: 16% hallucination rate → Their o3 model: 33% → Their newest o4-mini: 48% Nearly half of what their latest model tells you could be invented. And it's getting worse as models get "smarter." Here's why this can't be fixed: Language models predict the next word based on probability. When they hit uncertainty, they don't pause. They don't flag it. They guess with total confidence. Because that's literally what they were trained to do. The researchers analyzed the 10 major AI benchmarks used to test these models. 9 out of 10 give the exact same score for saying "I don't know" as for getting it completely wrong: zero points. The entire testing system punishes honesty and rewards confident guessing. So the AI learned the optimal strategy: always answer. Never show doubt. Sound certain even when making it up. OpenAI's proposed solution? Train models to say "I don't know" when uncertain. The problem? Their own math shows this would leave roughly 30% of questions unanswered. Imagine getting "I'm not confident enough to respond" three times out of ten. Users would abandon the product overnight. The fix exists. But it kills usability. This isn't just OpenAI's problem. DeepMind and Tsinghua University reached identical conclusions working separately. Three elite AI labs. Independent research. Same result: this is permanent. Every time you get an answer from any LLM, you're not getting facts. You're getting the most statistically probable next words from a system that's been rewarded for never admitting when it's guessing. Is this real information, or just a confident hallucination? You can't know. And neither can the AI.

super cool how searching for anything niche now takes over 10x longer because every goddamn search algorithm presumes it knows better than you, and that you must REALLY be looking for something more conventional

Look at these numbers. AI summaries are fuelling a news holocaust.