🫠 Startup Finds Way to Make AI Job Interviews Even Worse
｢ Adding insult to injury, the email's link to Alex's "mock interview" service led to a blank white page plastered with the words "coming soon." ｣
👩🏻🏫 Teachers got mad about a cheat button in Chrome. Now Google’s pausing it - The Washington Post
｢ A student taking an online quiz sees a button appear in their Chrome browser: “homework help.” Soon, Google’s artificial intelligence has read the question on-screen and suggests “choice B” as the answer ｣
This is funny
Mark Zuckerberg Demos New Facebook AI And It Couldn’t Have Gone Worse
https://kotaku.com/meta-ai-mark-zuckerberg-korean-steak-sauce-facebook-2000626808
Folks, it's happened. AI has become *so* smart that it knows how to intentionally sabotage a live product demos by Mark Zuckerberg and leave him helplessly flailing...
Joachim
boosted
@brianmerchant makes the shrewd point that #artists are losing jobs to #AI not because AI is that good, but because many companies that employ artists don't need work to be that good.
https://www.bloodinthemachine.com/p/artists-are-losing-work-wages-and
The same point applies far beyond the arts and we see examples every day. Poor writing in popular journalism. Poor decisions by insurance companies and other institutions. Poor attempts to grab your attention and influence you.
Merchant himself generalizes the point beyond AI and art: "There is a clear parallel to the Luddites here, who were skilled technicians and clothmakers who weren’t worried about technology surpassing them, but the way factory owners used it to make cheaper, lower-quality goods that drove down prices."
Greg Lloyd
boosted
Greg Lloyd
boosted
This also touches upon the age-old distinction between being smart and knowing facts. AI in the older definition of the word is about being smart, not (seemingly) knowing facts. And machine learning technology is not smart. But people with a naïve attitude to knowledge are really impressed when the LLM a) seems to understand what they say, b) always has an answer.
Greg Lloyd
boosted
It's not AI. It's a scam.
"under appropriate confidence thresholds, AI systems would naturally express uncertainty rather than guess. So this would lead to fewer hallucinations. The problem is what it would do to user experience.
Consider the implications if ChatGPT started saying "I don't know" to even 30% of queries ... Users accustomed to receiving confident answers to virtually any question would likely abandon such systems rapidly."
😄 😄 😄
@brianmerchant makes the shrewd point that #artists are losing jobs to #AI not because AI is that good, but because many companies that employ artists don't need work to be that good.
https://www.bloodinthemachine.com/p/artists-are-losing-work-wages-and
The same point applies far beyond the arts and we see examples every day. Poor writing in popular journalism. Poor decisions by insurance companies and other institutions. Poor attempts to grab your attention and influence you.
Merchant himself generalizes the point beyond AI and art: "There is a clear parallel to the Luddites here, who were skilled technicians and clothmakers who weren’t worried about technology surpassing them, but the way factory owners used it to make cheaper, lower-quality goods that drove down prices."
Ed Summers
boosted
Ice cold call to arms from the editors of n+1:
"AI-made material is itself a waste product: flimsy, shoddy, disposable, a single-use plastic of the mind"
🤜💀🤛
https://www.nplusonemag.com/issue-51/the-intellectual-situation/large-language-muddle/
It's not AI. It's a scam.
"under appropriate confidence thresholds, AI systems would naturally express uncertainty rather than guess. So this would lead to fewer hallucinations. The problem is what it would do to user experience.
Consider the implications if ChatGPT started saying "I don't know" to even 30% of queries ... Users accustomed to receiving confident answers to virtually any question would likely abandon such systems rapidly."
😄 😄 😄
This also touches upon the age-old distinction between being smart and knowing facts. AI in the older definition of the word is about being smart, not (seemingly) knowing facts. And machine learning technology is not smart. But people with a naïve attitude to knowledge are really impressed when the LLM a) seems to understand what they say, b) always has an answer.
It's not AI. It's a scam.
"under appropriate confidence thresholds, AI systems would naturally express uncertainty rather than guess. So this would lead to fewer hallucinations. The problem is what it would do to user experience.
Consider the implications if ChatGPT started saying "I don't know" to even 30% of queries ... Users accustomed to receiving confident answers to virtually any question would likely abandon such systems rapidly."
😄 😄 😄
Aral Balkan
boosted
I was thinking: the US AI industry (aka USAI-Inc) is like a new Empire colonisation. Like the East India Company. Or Belt & Road. Or others. Now its USAI. Many territories and sectors are being extorted and occupied (esp UK & UKHE). Then yesterday I saw that Berkman Klein had the same thought. They even called it a Raj. I felt vindicated, that I am not alone in what I think.
I was thinking: the US AI industry (aka USAI-Inc) is like a new Empire colonisation. Like the East India Company. Or Belt & Road. Or others. Now its USAI. Many territories and sectors are being extorted and occupied (esp UK & UKHE). Then yesterday I saw that Berkman Klein had the same thought. They even called it a Raj. I felt vindicated, that I am not alone in what I think.
jbz
boosted
👊 Patients Furious at Therapists Secretly Using AI
"He was taking what I was saying and putting it into ChatGPT."
Ice cold call to arms from the editors of n+1:
"AI-made material is itself a waste product: flimsy, shoddy, disposable, a single-use plastic of the mind"
🤜💀🤛
https://www.nplusonemag.com/issue-51/the-intellectual-situation/large-language-muddle/
China’s cyberspace regulator has instructed companies including Alibaba to halt orders for Nvidia’s RTX Pro 6000D, a semiconductor for workstations that can be repurposed for artificial-intelligence applications. https://www.japantimes.co.jp/business/2025/09/18/tech/china-nvidia-ai-chip/?utm_medium=Social&utm_source=mastodon #business #tech #china #nvidia #chipmakers #jensenhuang #ai #huawei
The challenge of maintaining curl / LWN
｢ Problematic email takes other forms as well. There is an increasing crowd of people who ask a large language model to "find a problem in curl, make it sound terrible", then send the result, which is never correct, to the project, thinking that they are somehow helping. Dealing with these useless problem reports takes an increasing amount of time ｣
New Python CLI Tool Catches MCP Server Issues Before Agents Do: https://thenewstack.io/new-python-cli-tool-catches-mcp-server-issues-before-agents-do/ via @TheNewStack & @sjvn
Microsoft Research MCP Interviewer needs work, but it's a step in the right direction for #AI agent Model Context Protocol interoperability.