The Big AI Joke

There are benefits in making people feel that computers are intelligent and devices are smart.

The first, and obvious one, is to imply that people are stupid and that, no matter what they claim, they will never be able to surpass what the machines can do. People will lose their job, acquiring skills through high education is just a waste of time and money. There is no need for learning many things, such as foreign languages or coding, among many other things. Any expertise is suspicious because it is being matched by AI agents. The goal, clearly, is to make people feel powerless.

The second benefit is the ability for giant companies to monopolize the construction of huge data processing centers as a way to show their dominance and to nuke any attempt to compete.

The third benefit is the fact that AI agents do not simply retrieve information that people are looking to get, they tell people how they should think. By returning a brief synthesis based on analyses of millions of sources, they tell us what answers are to be expected, be it for a class exam or to please a manager or an audience. The results reflect a statistical average response, without often any indication of where the conclusions come from. Using these agents require a blind act of faith into their super-powered assertions. The companies which are providing the AI agents have therefore all the latitude to manipulate the results to fit their purposes, and it opens the door to the ability for dictatorial regimes to spread political propaganda or lies. It is the step after fake news: it is information which is fabricated to serve certain interests. The ability for other information sources to provide meaningful information is clearly diminished.

It is possible to react to these threats.

1) It is possible to debunk the assertion that computers are smarter than humans. The work done by neuroscientists on human intelligence and the brain can be compared to the claims on how AI systems are "intelligent". Clearly, AI can provide a trove a useful information that is used in many fields, including biological research, but the ability to retrieve data that is stored does not equate with human intelligence. These are two different things.

2) The reason why data centers are huge is because a lot of information that is stored is trash. Companies that produce data are entirely capable to organize and present their data in a way that makes them accessible to their consumers. Once the data is vetted and properly presented, it is much easier to retrieve and it makes the need for huge data centers practically disappear.

3) Reliable data need to be sourced. There is a need for adding accountability to the information systems. The publishing system as we know it is based on printed information. What we need to do is to invent a new paradigm of publishing that is on par with the online distribution of information. This requires innovate legal work, the involvement of lawmakers to ensure that the foundational principles of democracy are met, and the creation of ethical and common-sense limitations on unregulated AI.