I, For One, Welcome Our New Self-Driving Overlords

13 Replies to “I, For One, Welcome Our New Self-Driving Overlords”

1. I wonder what they do with all the data. We have advertisers sucking up data 24/7 on every little thing to better reach consumers. Yet all the men in commercials (that aren’t retards) are black.

   1. Have you googled “panopticon”?

      1. Ah, yes, the brainchild of Jeremy Bentham, father of utilitarianism.
         The panopticon concept was insidious. You never know if you’re being watched, but you may be being watched, so behave, prisoner.
         As for Utilitarianism, the greatest good for the greatest number…right?

         1. And Fascism, the greatest good for the irrational number… Marxism, the greatest good for the imaginary number… Catholicism, the greatest good for the cardinal number…

   2. “ all the men in commercials (that aren’t retards) are black. “

      So true. Mass media has a burning hot hatred of white people. God forbid their commie puke masters ever gain full control. The Holocaust will look like a walk in the park.

2. It is gonna be all right. Amazon needs more warehouses.

3. All data centers are evil. The internet is a mature tech that needs proper maintenance.

4. Let’s just be clear here. AI might decide it wants us gone. Easiest method? AI starts a nuclear war. What might give AI pause about doing this? The data centres that the AI lives in would be destroyed by the nuclear war. What are we doing now? Planning data centres in space. Brilliant.

   RNrn

   1. AI data centres can be smashed with the proper hammers and enough willingness.

5. There are four leaders of AI. The last person who looked behind the curtain found one had no heart, one had no brain, one had no courage and one flew away in a hot air balloon. Then they woke up.

6. Arrogant Incompetence has no need for data centres.
   Other than as a cloak,The COMPUTER says..
   Just as the good book used to be used.
   Any “facts” required to promote the sacred narrative will be manufactured as required.
   just as “concensus” is manufactured via push polls.

7. Of the commenters here, I’m probably in the upper 10% of crazy, utopian, pro-AI. Meaning I’m somewhat positive on the tech, but really concerned about its social and economic impacts. Even if this individual is wrong (and there certainly are a lot of ghost data centres for lack of power) there are the following fairly obvious problems:
   – financing, to my untutored eye, resembles a circular bubble with pumped up GPU stocks and loans and equity flowing from the likes of NVidia to AI companies such as OpenAI, to data centre folks, back to the likes of NVidia to buy GPUs.
   – people are buying GPUs that rot at speed. Allegedly, Nvidia’s Blackwell platform of 2025/26 is eviscerated by the Vera Rubin platform due in Q3 which, if you believe the hypesters, reduces energy inferencing (models’ output) costs by 90%.
   – these data centres play havoc with local power stability (a glitch in power? Spin up the generators and drop off the grid, and a massive sudden load-side drop in demand. Hello voltage spikes, frequency spikes and a risk of cascading failure. All while nice smart lefties are urging unreliable solar and wind)

   In many cases, providers appear to be charging prices that don’t even seem to cover the electricity. How are they going to compete against newer entrants armed with substantial cost reductions when their capital runs out? (Yes, I get that they can sell your data, but more and more of what’s being pumped into AI is itself AI slop, which will lead to model collapse if used too much for training.)

   And just to make things really interesting, the Chinese are releasing open-weight models that look to be 80-95% as good (depending on what metrics you pick) and cost five to ten cents on the dollar compared to the top American models.

   A lot of people are going to lose a lot of money.

   1. Having lived through two previous super-hype => bust cycles in this industry, I assume the same thing will happen here: a lot of companies with no actual business model will go bankrupt once the seed capital runs out, and a small number of companies that figured out how to make something economical out of the tech will survive, and light the path for everyone who follows. Meanwhile, a lot of otherwise stable companies whose business model was leasing things to the first group will go under when they renege on all their contracts.

      It’s that last part that worries me. The Web 1.0 bust took out Nortel Networks and the Canadian tech industry never recovered. And a lot of very large companies have invested very stupidly in LLMs.