The solution for this is usually counter training. Granted my experience is on the opposite end training ai vision systems to id real objects.

So you train up your detector ai on hand tagged images. When it gets good you use it to train a generator ai until the generator is good at fooling the detector.

Then you train the detector on new tagged real data and the new ai generated data. Once it’s good at detection again you train the generator ai on the new detector.

Repeate several times and you usually get a solid detector and a good generator as a side effect.

The thing is you need new real human tagged data for each new generation. None of the companies want to generate new human tagged data sets as it’s expensive.

They license the anime so they do pay the studios that create anime. I know for a fact that anime studios factor in the ability to license shows in the decision on what to produce and budget.

The company had always been run by engineers that came up from chip fab. Then they fired both the CEO and the head of fab for sexual harassment.

Then they make the CFO with a MBA the new CEO. A year or two latter and chip design is having problems and fab is falling behind.

The Intel CEO had always come from engineering fab. This kept the high level decisions made by somebody who understood the product and how it was made.

Then CEO and the head of fab was caught sexually harassing employees. They were both shown the door. So no CEO and the guy who was next in line were gone. They needed the number 2 in fab to take over fab to keep production up.

So the board decided to make the CFO the new CEO. A guy who had a MBA was running a chip company that had only been run by engines.

Profits went up for a while but then Intel struggled to maintain innovation and properly upgraded fab and chip design. Add the increasing skill of rivals and a increase in importance in chips other then server and desktop. which were the only areas Intel was king. It’s a recipe for failure.

Algae doesn’t capture it for long. Trees do it for longer but not long enough to be more then a speed bump. Unless we start dumping algae and trees into giant pits and sealing them up three is no long term carbon capture.

Prisencolinensinainciusol an Italian song that is complete gibberish but made to sound like an English language song. That’s what AI is right now.

https://www.youtube.com/watch?v=RObuKTeHoxo

Berkshire Hathaway owns Geico the car insurance company. In one of his annual letters Buffett said that autonomous cars are going to be great for humanity and bad for insurance companies.

“If [self-driving cars] prove successful and reduce accidents dramatically, it will be very good for society and very bad for auto insurers.”

Actuaries are by definition bad at assessing new risk. But as data get collected they quickly adjust to it. There are a lot of cars so if driverless cars become even a few percent of cars on the road they will quickly be able to build good actuarial tables.

We have set up the world to be very usable and traversable by humans. A robot in that form would be able to go and do most things. Think of all the changes to places that needed to be made as part of the ADA. And those were for humans with relatively small changes such as wheels instead of legs.

It’s also a win win design. Shade from the panels reduces evaporation in the canals and the water helps cool the panels which improves their efficiency.

The ai generated portion does not have copy right protection. This also applies within an image. So for instance in an ai generated building image with a human created character in front. People couldn’t copy the character but could use the background.