941
Microsoft CEO Admits That AI Is Generating Basically No Value (ca.finance.yahoo.com)
submitted 2 years ago by to c/technology@lemmy.world
123 comments fedilink hide all child comments
 

"The real benchmark is: the world growing at 10 percent," he added. "Suddenly productivity goes up and the economy is growing at a faster rate. When that happens, we'll be fine as an industry."

Needless to say, we haven't seen anything like that yet. OpenAI's top AI agent — the tech that people like OpenAI CEO Sam Altman say is poised to upend the economy — still moves at a snail's pace and requires constant supervision.

you are viewing a single comment's thread
view the rest of the comments
[–] 20 points 2 years ago (2 children)

I think AI is generating negative value ... the huge power usage is akin to speculative blockchain currencies. Barring some biochemistry and other very, very specialized uses it hasn't given anything other than, as you've said, plain-language search (with bonus hallucination bullshit, yay!) ... snake oil, indeed.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 9 points 2 years ago (1 child)

    Its a little more complicated than that I think. LLMs and AI is not remotely the same with very different use cases.

    I believe in AI for sure in some fields, but I understand the skeptics around LLMs.

    But the difference AI is already doing in the medical industry and hospitals is no joke. X-ray scannings and early detection of severe illness is the one being used specifically today, and will save thounsands of lives and millions of dollars / euros.

    My point is, its not that black and white.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 7 points 2 years ago

    On this topic, the vast majority of people seem to think that AI means the free tier of ChatGPT.

    AI isn't a magical computer demon that can grant all of your wishes, but that doesn't mean that it is worthless.

    For example, Alphafold essentially solved protein folding and diffusion models built on that discovery let us generate novel proteins with specific properties with the same ease as we can make a picture of an astronaut on a horse.

    Image classification is massively useful in manufacturing. Instead of custom designed programs purpose built for each client ($$$), you can find tune existing models with generic tools using labor that doesn't need to be a software engineer.

    Robotics is another field. The amount of work required for humans to design and code their control systems was enormous. Now you can use standard models, give them arbitrary limbs and configurations and train them in simulated environments. This massively cuts down on the amount of engineering work ($$$) required.

  • permalink
  • fedilink
  • source
  • parent