home.social

#thinkingmachines — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #thinkingmachines, aggregated by home.social.

  1. The eInk faceplate #Valve demoed for the #SteamMachine makes me wonder if there's some sort of power/data port hidden behind it. I'd love to turn one into a mini-#ThinkingMachines #CM2 complete with an array of blinkenlights representing system load.

  2. #ThinkingMachines Lab, founded by #MiraMurati, is working on making #AImodels more #consistent by addressing the randomness in their responses. The lab’s research suggests that controlling the way GPU kernels are orchestrated during inference processing can lead to more #deterministic AI models. techcrunch.com/2025/09/10/thin #tech #media #news

  3. #ThinkingMachines Lab, founded by #MiraMurati, is working on making #AImodels more #consistent by addressing the randomness in their responses. The lab’s research suggests that controlling the way GPU kernels are orchestrated during inference processing can lead to more #deterministic AI models. techcrunch.com/2025/09/10/thin #tech #media #news

  4. #ThinkingMachines Lab, founded by #MiraMurati, is working on making #AImodels more #consistent by addressing the randomness in their responses. The lab’s research suggests that controlling the way GPU kernels are orchestrated during inference processing can lead to more #deterministic AI models. techcrunch.com/2025/09/10/thin #tech #media #news

  5. #ThinkingMachines Lab, founded by #MiraMurati, is working on making #AImodels more #consistent by addressing the randomness in their responses. The lab’s research suggests that controlling the way GPU kernels are orchestrated during inference processing can lead to more #deterministic AI models. techcrunch.com/2025/09/10/thin #tech #media #news

  6. #ThinkingMachines Lab, founded by #MiraMurati, is working on making #AImodels more #consistent by addressing the randomness in their responses. The lab’s research suggests that controlling the way GPU kernels are orchestrated during inference processing can lead to more #deterministic AI models. techcrunch.com/2025/09/10/thin #tech #media #news

  7. Some of you probably remember WAIS. An Internet search engine from the Gopher era (more specific: a search protocol). TIL: Thinking Machines (those with the massive red blinkenlights servers) were the main driver and developer of WAIS: en.wikipedia.org/wiki/Wide_are #vintagecomputing #techhistory #wais #thinkingmachines #connectionmachine