To gather training data directly from peoples behavior for free. To build a model of you. Then to exploit that model. They have run out of training data from the web, so they are going to collect it directly from people. The agents may eventually do some useful things for users, but their main purpose is to train on you.

and it will not stop financial theft either, it just makes it easier for you to notice it.

or they are not fit for their jobs.

At this point, after decades of research, there appears to be no solution for computer security. No solution to ransomware, no solution to zero-days, no solution to spying, no solution to infrastructure integrity, no solution to social engineering. So the end game must be that there will be no privacy for anyone, companies, CEO's, Presidents, law enforcement, citizens, everyone. Maybe we have to go back to training people to be good citizens, or give up computers and networks.

If there are more jobs posted than actually exist then it looks like anyone who wants a job can get one, but if the jobs are fake listings then decisions about the national economy will be based on false information.

It takes much longer to bring other forms of power online. So all those AI investments are doomed.

This is how you achieve AGI and reduce the cost and power requirements of AI, all at the same time. It would also work for self driving vehicles, humanoid robots, business logic, autonomous weapons, and most jobs. It could even replace politicians.

I was involved in some research on a stress detecting algorithm for a smartwatch and this is the correct take on what is being measured. There are everyday stressors and there are long term stressors. Long term stress harms health. Short term stress is normal. The main measurement for stress is heart rate variability (HRV), not just heart rate. Your heart when you are active and excited runs at a steady beat. When you are relaxing it delays beats sometimes. That delay is HRV. If your HRV is low all the time your heart is maintaining a steady beat like you are about to run a race or run from danger, which means the heart never gets to relax a little. If your HRV is high the heart is getting chances to take a break. We validated our algorithms using things like plunging a hand into ice cold water or giving a public speech, which are well known to cause stress levels to rise. The gold standard would be to measure cortisol levels in the blood or saliva, but that's expensive. Most smartwatches report a measure derived from HRV in the moment. It's you that has to look at this data over a long time period and determine if you have consistent continuous high levels of stress (low HRV), the smartwatch usually does not provide much support for that kind of observation. They may compute an average, but if your stress is always high the average may not tell you much if you are looking for changes in stress (people fixate on the short term changes, not the long term changes or steady state). Optical sensors are good enough now that they can do a fairly good job of measuring HRV by averaging over several heartbeats, if you are not moving and the sensor has good skin contact. The accelerometer can be used to only take readings when you are still. How each companies algorithm works is proprietary, but the basics of a good algorithm are widely known. I'd say this study was kind of useless, typical of psychiatry, they asked people for their opinion, which is notoriously unreliable, instead of measuring cortisol.

They want to be recognized as sentient and indigenous so they can hire pro bono lawyers and sue the ground dwellers for polluting their nation and stealing all the best food.

How many more assumptions are out there waiting to be retested and found to be wrong? I suspect that quite a few scientists could make a career of testing assumptions and make great progress on a number of long standing medical issues. Actually, this doesn't just apply to medicine, it applies to any assumption ever made. Those old founders of science had a lot of hubris that no one has challenged for many decades.

You pay people less, then they can not afford to buy products. Corporations have been destroying their own markets for decades and now it is affecting their bottom line.

Workers can use AI on their own to automate their own jobs and give themselves more time. They can use that time to get more jobs which they can then also automate. Soon it may be common to have ten jobs, mostly handled by your own AI. Then working for any one company matters less, people may quit some jobs if they feel like going on vacation. The AI sword cuts both ways, workers can also use it to automate their work too. As AI improves use your personal AI to start a company run by the AI. No more bosses.

Accepts remote commands, controls all apps, reports everything to a centralized server.

If these CEOs are bailing because they see problems ahead then divesting from their companies would seem to be a winning strategy.

Texas is leading the country in renewables, will Trump cancel all their new projects and turn off their already built projects?