Fuck AI

Meta is one of several tech companies vying for a nuclear boost.

Meta has faced a setback in its plan to build data centers run on nuclear power. The FT reports that CEO Mark Zuckerberg told staff last week that the land it was planning to build a new data center on was discovered to be the home of a rare bee species, which would have complicated the building process.

OpenAI’s Whisper tool may add fake text to medical transcripts, investigation finds.

Research paper.

• A new OpenAI study using their SimpleQA benchmark shows that even the most advanced AI language models fail more often than they succeed when answering factual questions, with OpenAI's best model achieving only a 42.7% success rate.
• The SimpleQA test contains 4,326 questions across science, politics, and art, with each question designed to have one clear correct answer. Anthropic's Claude models performed worse than OpenAI's, but smaller Claude models more often declined to answer when uncertain (which is good!).
• The study also shows that AI models significantly overestimate their capabilities, consistently giving inflated confidence scores. OpenAI has made SimpleQA publicly available to support the development of more reliable language models.

This article is talking about phishing websites made by scammers with obvious signs that it was made by LLMs.

I thought it might be interesting here.

Meta is “working with the public sector to adopt Llama across the US government,” according to CEO Mark Zuckerberg.

The comment, made during his opening remarks for Meta’s Q3 earnings call on Wednesday, raises a lot of important questions: Exactly which parts of the government will use Meta’s AI models? What will the AI be used for? Will there be any kind of military-specific applications of Llama? Is Meta getting paid for any of this?

When I asked Meta to elaborate, spokesperson Faith Eischen told me via email that “we’ve partnered with the US State Department to see how Llama could help address different challenges — from expanding access to safe water and reliable electricity, to helping support small businesses.” She also said the company has “been in touch with the Department of Education to learn how Llama could help make the financial aid process more user friendly for students and are in discussions with others about how Llama could be utilized to benefit the government.”

She added that there was “no payment involved” in these partnerships.

yeah fck them, for now until the government relies on their AI

These are better than those weird videos.

Somehow it missed the massive forest fire this summer that destroyed much of the park and the town ... until it was reminded.

Remind me why anybody takes this tech seriously?

    (archived link)

Google could preview its own take on Rabbit’s large action model concept as soon as December, reports The Information. “Project Jarvis,” as it’s reportedly codenamed, would carry tasks out for users, including “gathering research, purchasing a product, or booking a flight,” according to three people the outlet spoke with who have direct knowledge of the project.

If a robot ever buys something on my behalf, I'm lawyering the fuck up.

cross-posted from: https://lemmy.world/post/21301373

Tech behemoth OpenAI has touted its artificial intelligence-powered transcription tool Whisper as having near “human level robustness and accuracy.”

But Whisper has a major flaw: It is prone to making up chunks of text or even entire sentences, according to interviews with more than a dozen software engineers, developers and academic researchers. Those experts said some of the invented text — known in the industry as hallucinations — can include racial commentary, violent rhetoric and even imagined medical treatments.

Experts said that such fabrications are problematic because Whisper is being used in a slew of industries worldwide to translate and transcribe interviews, generate text in popular consumer technologies and create subtitles for videos.

More concerning, they said, is a rush by medical centers to utilize Whisper-based tools to transcribe patients’ consultations with doctors, despite OpenAI’ s warnings that the tool should not be used in “high-risk domains.”

Is OpenAI breaking U.S. copyright law? A former employee of the company says yes.

A former researcher at the OpenAI has come out against the company’s business model, writing, in a personal blog, that he believes the company is not complying with U.S. copyright law. That makes him one of a growing chorus of voices that sees the tech giant’s data-hoovering business as based on shaky (if not plainly illegitimate) legal ground.

Guess it's time to cut off OpenAI's internet service. That's how it works, right, Copyright Cartel??

Imagine walking into your office to find that your company just hired thousands of new employees overnight – except they're not human. That's exactly what Microsoft has made possible with its groundbreaking announcement of autonomous AI agents, marking a fundamental shift in how businesses will operate in the coming years.

Unlike traditional AI assistants that simply respond to commands, these new autonomous agents can independently initiate and complete complex business tasks. Through Microsoft's Copilot Studio, organizations can create AI employees that handle everything from qualifying sales leads to managing supplier communications. These agents don't just follow predetermined scripts – they analyze situations, make decisions, and take action without human intervention.

Early adopters are seeing remarkable results.

McKinsey & Company implemented an AI agent that reduced client onboarding lead times by 90% and cut administrative work by 30%. At Pets at Home, the UK's leading pet care business, an AI agent handling profit protection cases is projected to deliver seven-figure annual savings. These aren't just incremental improvements – they represent a fundamental transformation in business operations.

M$ has invested a lot of money in AI in various countries. They will take contracts from the government, as a result there will be less new hires for government jobs.

https://www.reuters.com/technology/microsoft-make-27-billion-cloud-ai-investments-brazil-2024-09-26/

https://www.apac-business.com/companies/market/microsoft-announces-ai-skilling-opportunities-for-2-5-million-people-in-the-asean-region-by-2025/

cross-posted from: https://lemmy.bestiver.se/post/67080

Comments