The Social Web (old posts, page 210)

RSS feed

macOS 26 May Not Support 2018 MacBook Pros, 2019 iMacs, or the iMac Pro

Apple's upcoming macOS 26 operating system may abandon support for several older Mac models, according to AppleInsider. The casualties will include 2018 MacBook Pro models, the 2020 Intel MacBook Air, the 2017 iMac Pro, and the 2018 Mac mini -- all currently the oldest machines compatible with macOS Sequoia, the report said, citing a source familiar with the matter. The 2019 MacBook Pro models and 2020 5K iMac models will retain compatibility with the new system, codenamed "Cheer," said AppleInsider.

Read more of this story at Slashdot.

HP Hastens China Exit as Tariffs Kick a Hole in its Profits

An anonymous reader shares a report: HP is close to ending production of North-America-bound products in China, after US tariffs kicked a hole in its quarterly profits. "A quarter ago, we shared that our goal was to have less than ten percent of the products in North America being shipped from China by September," HP president and CEO Enrique Lores told investors on the company's Q2 2025 earnings call. "We have accelerated that and we share that now almost no products will be coming from China sold in the US by June. It's a very significant acceleration of the plan that we have." "We accelerated the shift of factories out from China into Southeast Asia, into Mexico to a certain extent in the US to mitigate the impact of the change," he added. Lores also revealed that HP has removed the US as a distribution hub for products sold in Canada or to Latin America. Doing so means HP doesn't have to pay tariffs.

Read more of this story at Slashdot.

Researchers Warn Against Treating AI Outputs as Human-Like Reasoning

Arizona State University researchers are pushing back [PDF] against the widespread practice of describing AI language models' intermediate text generation as "reasoning" or "thinking," arguing this anthropomorphization creates dangerous misconceptions about how these systems actually work. The research team, led by Subbarao Kambhampati, examined recent "reasoning" models like DeepSeek's R1, which generate lengthy intermediate token sequences before providing final answers to complex problems. Though these models show improved performance and their intermediate outputs often resemble human scratch work, the researchers found little evidence that these tokens represent genuine reasoning processes. Crucially, the analysis also revealed that models trained on incorrect or semantically meaningless intermediate traces can still maintain or even improve performance compared to those trained on correct reasoning steps. The researchers tested this by training models on deliberately corrupted algorithmic traces and found sustained improvements despite the semantic noise. The paper warns that treating these intermediate outputs as interpretable reasoning traces engenders false confidence in AI capabilities and may mislead both researchers and users about the systems' actual problem-solving mechanisms.

Read more of this story at Slashdot.