Several major AI companies — including OpenAI, Google, Microsoft and Anthropic — have backed a shared signal that lets websites block their content from being used to train AI models. The proposal, called __usage_ai-training, would be added to a site's robots.txt file and was drafted by Anthropic's Kevin Kelley on Oct 30. Meta has signaled support but hasn't announced when it will adopt the standard. The measure is voluntary and doesn't resolve ongoing disputes over copyright or payment for training data.
Meanwhile, Google achieved a notable milestone in space-based AI computing. The company announced it successfully ran an AI model on a data-center-class processor aboard a satellite as part of Project Suncatcher. Engineers celebrated the test as a key step toward exploring whether orbital data centers could one day support AI workloads.
In the job market, a growing number of companies are rethinking AI-driven layoffs. Microsoft, Barclays Bank and several supermarket chains cut workers citing AI efficiency, only to bring some back later — a trend dubbed the AI boomerang. Ford hired 350 engineers after finding that AI quality tools underperformed. Experts say AI still struggles to replace institutional knowledge, creativity and critical thinking that humans bring to the table.
Public sentiment toward AI infrastructure remains mixed. A Gallup poll found that seven in ten Americans oppose having a data centre built near them. Voters in 38 local jurisdictions will weigh at least 41 measures on data centres on Tuesday, with 21 of those measures in Ohio where both Democrats and Republicans stand opposed. These facilities power services like ChatGPT and Claude.
On the technical front, scientists have found new ways to improve how AI chatbots handle text. A method called byteification converts models to read at the letter level rather than whole word chunks, improving spelling without hurting performance. Three models — Qwen3, Llama 3 and Olmo 3 — were converted into Bwen, Blama and Bolmo using less than 1% of the training data typically required.
In healthcare, Sheba Medical Center in Israel plans to become the first hospital outside the US to implement the latest OpenAI capabilities across all its wards by 2027. Sheba Impact CEO Avner Halperin is leading the effort to deliver more accurate and efficient treatment through AI tools.
Kazakhstan is charting its own course in the AI economy, seeking to benefit from technology without fully aligning with either the US or China. Chinese car brands like BYD and Changan are becoming increasingly common in the capital Astana as the Central Asian nation navigates the growing tech divide between superpowers.
On the question of trust, experts emphasize that AI chatbots still fall short when it comes to evaluating information quality. The CRAAP test — which checks for Currency, Relevance, Authority, Accuracy and Purpose — requires the kind of contextual judgment and bias detection that current AI tools lack. Human critical thinking remains essential for sorting reliable sources from unreliable ones.
In film, the indie movie Anna by the Window, co-directed by Myles Flusser and James Hyland, explores how technology affects personal relationships. The story follows an AI startup founder and his family, developing more as a character-driven drama than a science fiction tale.
Key Takeaways
- <ul><li>OpenAI, Google, Microsoft and Anthropic support a common robots.txt signal called __usage_ai-training that lets websites opt out of AI model training.</li><li>Meta supports the standard but has not announced when it will implement it.</li><li>The proposal was drafted by Anthropic's Kevin Kelley on Oct 30 and remains voluntary.</li><li>Google successfully ran an AI model on a satellite processor as part of Project Suncatcher.</li><li>Companies like Microsoft, Barclays and Ford are rehiring workers after AI tools underperformed — a trend called the AI boomerang.</li><li>Ford hired 350 engineers after AI quality tools failed to meet expectations.</li><li>A Gallup poll found seven in ten Americans oppose data centres near their homes, with 41 measures on ballots in 38 jurisdictions.</li><li>Byteification converts AI models to read at the letter level, improving spelling using less than 1% of normal training data.</li><li>Sheba Medical Center in Israel will implement OpenAI tools across all wards by 2027, becoming the first non-US hospital to do so.</li><li>Kazakhstan is building an AI economy while maintaining ties with both the US and China.</li></ul>
New standard helps publishers opt out of AI training
Major AI companies including OpenAI, Google, Microsoft and Anthropic support a new shared token for websites. The token would let publishers opt out of having their content used to train AI models. The signal would be added to a site's robots.txt file. Meta has signaled support but has not said when it will follow the standard.
AI companies back one signal to reject model training
Google, Microsoft, OpenAI and Anthropic support a common robots.txt control called __usage_ai-training. The signal lets websites state that their content should not be used to train AI models. The proposal was drafted by Anthropic's Kevin Kelley on Oct 30. It remains voluntary and does not settle copyright or payment disputes.
New film explores human connection in the AI age
The indie film Anna by the Window is co-directed by Myles Flusser and James Hyland. It follows an AI startup founder and his family. The story feels less like science fiction as it develops. The film examines how technology affects personal relationships.
Data centre votes could shape US midterms
Voters in 38 local jurisdictions will consider at least 41 measures about data centres on Tuesday. These buildings power AI services like ChatGPT and Claude. A Gallup poll found seven in ten Americans oppose a data centre being built near them. Twenty-one of the measures are in Ohio where Democrats and Republicans oppose them together.
What to know when companies rehire after AI job cuts
Companies are calling the trend of rehiring workers the AI boomerang. Microsoft Barclays Bank and supermarket chains have cut jobs citing AI then later brought workers back. Experts say AI cannot replace institutional knowledge creativity and critical thinking. Ford hired 350 engineers after AI quality tools underperformed. Workers should focus on their unique human strengths.
Kazakhstan stays neutral while pursuing AI economy
Kazakhstan is working to build an AI economy without choosing between the US and China. The Central Asian nation is navigating a growing tech divide between the two superpowers. Chinese car brands like BYD and Changan are becoming common in the capital Astana. The country wants to benefit from AI technology while keeping ties with both powers.
Google tests AI model on satellite processor
Google announced it successfully ran an AI model on a data-center-class processor aboard a satellite. This milestone is part of Project Suncatcher, which explores whether AI computing can work in space. The test shows early progress toward orbital data centers. Google engineers celebrated the achievement as a key step forward.
AI chatbots cannot replace human critical thinking
The CRAAP test is a tool for checking if sources are trustworthy, standing for Currency, Relevance, Authority, Accuracy, and Purpose. AI chatbots lack the ability to apply this test because they cannot spot biases or understand context. Human critical thinking remains essential in a world filled with information. AI tools still need people to judge the quality of information.
New method helps AI chatbots read individual letters
Scientists converted AI models to read text at the letter level instead of whole word chunks. The process called byteification improved spelling skills without hurting performance. Three models named Qwen3, Llama 3 and Olmo 3 were converted into Bwen, Blama and Bolmo. The conversion used less than 1% of the training data normally needed.
Sheba hospital to use OpenAI tools by 2027
Sheba Medical Center in Israel plans to implement AI across all its wards by 2027. It is the first hospital outside the US to get the latest capabilities of OpenAI. Sheba Impact CEO Avner Halperin leads the effort to make it a leading AI hospital. The hospital aims to provide more accurate and efficient treatment through AI.
Sources
- How a new standard could make it easier for publishers to opt out of AI training
- NetNewsLedger - Major AI companies back one website signal for rejecting model training
- New film 'Anna by the Window' explores human connection in the AI age
- Tech 24 - Data centres on the ballot: could the AI infrastructure shape US midterms?
- What to consider if a company that made AI-related job cuts wants you back
- US or China? Why Kazakhstan refuses to pick sides as it eyes AI revolution
- Google runs AI model on satellite chip, advancing orbital data center ambitions
- Mastering the CRAAP test: Why AI chatbots can't replace human critical thinking
- Scientists give AI chatbots a new way to read, with surprising results
- "We're the first hospital outside the US that gets the latest capabilities of OpenAI"
Comments
Please log in to post a comment.