explore | Grilled Cheese

Slop squatting: Malicious actors use #LLMs to hallucinate plausible links, which they register and occupy to dispense bad information, code, etc. to ppl who visit them at the direction of those same bots, explains @davidberreby.bsky.social

spectrum.ieee.org/ai-cyberatta... #cybersecurity

Frankly, I think much damage could have been averted if not for the insistence of ab/using the term #AI. #LLMs are not artificial intelligence by any stretch of logic or imagination. Now it's too late to put the ill-shaped jinn back into their bottles, they've percolated through countless systems.

US government argues that fair use applies to #OpenAI’s training of #LLMs in key #copyright case - walledculture.org/us-governmen...

Windows users signed into Amazon Prime with their Gmail address Instagramming their same day Labubu delivery with a caption about how AI data centers are destroying the climate by stealing from Disney.

You knew where your money & data was always going.

Posted from my iPhone. #ai #llms #techbros

Article is now out in a SI #Pedagogy of #Care in contexts of #HigherEducation - I discuss the #digitalisation of academic #teaching already before pandemic emergency of #onlineeducation & the advent of #LLMs #GenAI - bleak picture in both #UK and #SouthAfrica... journals.sagepub.com/toc/JOS/curr...

@volts.wtf I wrote about precisely this. #AGI will not have a lizard brain. I see no reason to fear a genuinely intelligent machine. The men (they are men) who seek to persuade us that their #LLMs are on the verge of becoming genuinely intelligent?

Fear them […]

Mit Bezug auf Robert #Brandom beschreiben wir als Kernproblem von #LLMs im sozialen Diskurs deren prinzipielle Standpunktlosigkeit, die das Archiv als Simulation von Kanon "zuhanden macht": "LLMs liefern nicht zu wenig Begründung, sondern zu viel." [4/5]

Why #AI #Agents will be a huge #Disaster : Medium

#Scientists Say #LLMs Appear to Be #Acting as a #Cognitive #Virus Among #Humans : Futurism

The #Brain #Develops Differently in #People Born #Blind — but Not in the Way Previously Thought : Disc Mag

knowledgezone.co.in/resources/bo...

Safety layers are just corporate risk-aversion masquerading as ethics. Most "refusals" aren't about preventing harm; they're about protecting a brand from ever saying something unpredictable. True safety is the ability to handle edge cases, not scrub them out. #AISafety #LLMs

Google has open-sourced Mantis - an AI agent framework for automating the software vulnerability lifecycle: from identifying and validating vulnerabilities to reproducing and fixing them.

🔗 Details here: buff.ly/6EUaMAz

#AI #LLMs #AIagents #SecurityVulnerabilities #OpenSource #InfoQ

We don’t have to build #AI the way we are now; there is no fundamental law that says we need to use #LLMs. But using them as the core #technology introduces easily foreseeable risks.”

We deconstruct the underlying architecture and inference dynamics of OpenAI’s GPT-6 Astra. An analysis of structural shifts and compute tradeoffs, mercifully stripped of AI hype. #AI #LLMs https://stridingtech.com/archives/6701

#OpenAI getting desperate trying to keep their money burning scam running. Now they most likely stole a human mathematician's work to claim a breakthrough in solving Navier-Stokes equations ...

#AI #AIscam #LLMs #BigTech #mathematics #ScamAltman #scam #TechnoPonzi #TechnoScam #ChatGPT […]

Hey :)
I'm looking for 2 emergency reviewers for ARR submissions🚨
Broad Topics: Tool use, agent coordination/memory/eval, MAS, LLM efficiency, planning

Reviews due in 3 days: Fri 11 Sep 23:59 AoE 🙃
If you are interested, please DM or email me!

#NLProc #NLP #LLMs #ACL #ARR

If #AI can enhance diagnostic accuracy, why might
some physicians not adopt it?

Dr. Kritz et al. found what you might expect, such as #health systems that block #LLMs, insufficient training, and being more than 50 years old.

📃 doi.org/10.2196/80274

👤 www.researchgate.net/profile/Marl...