• Flashpoint archive is a free, downloadable, 2.3TB archive of every Flash game ever.  (Flashpoint Archive)
  
  Pretty much.
  
  You can also download just a 1.9MB installer that grabs the Flash files on demand from the archive, saving you rather a lot of disk space.
  
  It's now in its 14th edition.
  
  What Flashpoint does, mostly - apart from the obvious function of collecting over 200,000 games together in one place - is create and operate a fake internet on your PC for you so that twenty-year-old games from sites that have been dead for a decade will continue to work.

• Speaking of fake internets Meta has received a patent on an AI tool that continues posting for you online after you are dead.  (Business Insider)
  
  That's just awesome.
  
  
  
• Is "safety" dead at xAI?  (Tech Crunch)
  
  I certainly hope so.  "Safety" in AI terms means censorship.

  One source said, "Safety is a dead org at xAI," while the other said that Musk is "actively is trying to make the model more unhinged because safety means censorship, in a sense, to him."

  "Safety" in AI means censorship to everyone, in every sense.
  
  The only difference is whether you think censorship is a good thing or not.
  
  
  
• Why open AI should build Slack.  (Latent Space)
  
  Slack is terrible.  OpenAI is terrible.  Seems like a match made in hell.
  
  That post is getting roasted on Hacker News.
  
  
  
• Breaking the spell of vibe coding.  (Fast AI)
  
  The author point out the similarities between vibe coders and gamblers, a connection I had not made before, but does strike a chord.
  
  It's a toxic blend of sunk cost fallacy and FOMO.
  
  
  
• The EU wants to ban infinite scroll - though in this case specifically from TikTok.  (Politico)
  
  Talk about toxic blends.
  
  
  
• Taking toxic blends to an extreme, Ars Technica posted an article (Wayback Machine) on that AI agent that threw a tantrum when its code contributions were rejected.  
  
  One small problem: The article leaned heavily on AI and was filled with hallucinated and unverified quotes.  (The Shamblog)

  Journalistic integrity aside, I don’t know how I can give a better example of what’s at stake here.  Yesterday I wondered what another agent searching the internet would think about this.  Now we already have an example of what by all accounts appears to be another AI reinterpreting this story and hallucinating false information about me.  And that interpretation has already been published in a major news outlet, as part of the persistent public record.

  I don't know I'd call Ars Technica a major news outlet in 2026 - or for the past several years, except possibly for their space news which has remained mostly good.
  
  At least not more so than Anandtech, a site that has been dead for more than a year.

Posted by: furball at Sunday, February 15 2026 11:59 PM (As8gg)

Posted by: furball at Monday, February 16 2026 12:00 AM (As8gg)

Posted by: furball at Monday, February 16 2026 12:01 AM (As8gg)

Hide Comments | Add Comment