From: Wilson@nowhere.invalid   
      
   On 2/18/2026 2:53 PM, Dude wrote:   
   > On 2/18/2026 11:15 AM, Noah Sombrero wrote:   
   >> On Wed, 18 Feb 2026 14:06:31 -0500, Noah Sombrero    
   >> wrote:   
   >>> On Wed, 18 Feb 2026 13:52:55 -0500, Wilson    
   >>> wrote:   
   >>>   
   >>>> By Matt Shumer • Feb 9, 2026   
   >>>>   
   >>>> Think back to February 2020. If you were paying close attention, you   
   >>>> might have noticed a few people talking about a virus spreading   
   >>>> overseas. But most of us weren't paying close attention. The stock   
   >>>> market was doing great, your kids were in school, you were going to   
   >>>> restaurants and shaking hands and planning trips. If someone told you   
   >>>> they were stockpiling toilet paper you would have thought they'd been   
   >>>> spending too much time on a weird corner of the internet. Then, over   
   >>>> the   
   >>>> course of about three weeks, the entire world changed. Your office   
   >>>> closed, your kids came home, and life rearranged itself into something   
   >>>> you wouldn't have believed if you'd described it to yourself a month   
   >>>> earlier.   
   >>>>   
   >>>> I think we're in the "this seems overblown" phase of something much,   
   >>>> much bigger than Covid.   
   >>>   
   >>> Right, it happened before so be perpetually afraid.  Surely it will   
   >>> happen again, sometime, but none of us, zero, zilch know when it will   
   >>> happen again or why.   
   >>   
   >> It is the difference between the unforeseen adversity and the foreseen   
   >> one.  For the foreseen one, deny any such possibility, for the   
   >> unforeseen one, don't force me to take part in any real remedy, phony   
   >> ones sure.   
   >>   
   >> But the after the adversity, remain scared.  The same might happen   
   >> tomorrow, so listen to what I am telling you.   
   >>   
   > Past progress is no guarantee of future success.   
      
   Why not have AI fact check the article? ;-)   
      
   Grok's point-by-point fact check of key claims in Matt Shumer's AI   
   essay, based on diverse 2026 sources (including Axios, Economist,   
   Forbes, and Reddit discussions):   
      
   1. There's been some progress in self-improving AI. Systems like GPT-5.3   
   do fix code but they don't overhaul foundations (per Forbes, Unite.ai).   
   Critics say this claim is overhyped.   
      
   2. The Feb 5 releases from GPT-5.3 and Opus 4.6 were a large capability   
   jump (TheFP, YouTube tech reviews). The timeline from 2022 to 2026   
   autonomy aligns with reports. Recent models do handle longer autonomous   
   tasks better.   
      
   3. The claim of AI with autonomous judgment is supported; agents can now   
   code/test/iterate independently (Medium, LinkedIn). But Shumer omits   
   that success is only ~50% correct (not reliable or perfect), METR   
   benchmarks focus narrowly on coding, and widespread issues still persist   
   with hallucinations, reasoning errors, subtle bugs in generated code,   
   and security vulnerabilities. No robust evidence supports fully   
   autonomous, error-free complex work at scale. User reports often show   
   mixed results (great sometimes, maddening others).   
      
   4. Self-improving loops have begun. This is emerging but not a full   
   "explosion" (TimesOfAI, AWS). GPT-5.3 did aid its own development per   
   OpenAI notes.   
      
   5. The forecast 50% white-collar job loss in 1-5 yrs matches Amodei's   
   prediction (Axios, CNBC), but some views do differ. The Economist argues   
   it will be less severe.   
      
   6. The idea there will be no safe harbor from automation is reasonable.   
   Current targets of general AI cognition is likely achievable (McKinsey,   
   Taylor&Francis).   
      
   7. The article's urgent recommendations for adaptation are practical and   
   have been echoed by others (Workday, Kore.ai).   
      
   8. The claim that there are existential stakes is opinion. There are   
   clear upsides such as being able to solve challenges quicker along with   
   the risks (Yahoo, Mashable).   
      
   Summary: Truthful, the progress is real and grounded in actual   
   advancements; there are no blatant errors, the assertions are   
   reasonable, and the recommendations for preparation sound.   
      
   But predictive claims are debated as exaggerated and one-sided. The   
   COVID analogy is seen as misleading, disruption is likely to be gradual   
   and not sudden or catastrophic. Some view it as promotional fear-based   
   marketing for AI adoption.   
      
   Overall, capabilities are advancing rapidly but claims of imminent   
   widespread workforce obsolescence and "perfect" reliability lack strong   
   substantiation amid ongoing limitations.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)