The glorious AI future

is not as glorious as you probably think.

Let’s outline the glorious AI future that our politicians are so enthusiastic about.

• In the near (how near?) future AI systems will really be intelligent.
• They will also be cheap and so become pervasive: everybody will have an AI assistant, or perhaps large numbers of them.
• This glorious future is worth almost any sacrifice: the carbon emissions from the vast machines we need to create it will somehow be solved by magic unspecified technology developed by the AIs of the future, and the sacrifice of all our intellectual property to the soul-harvesting parasites splendid AI will enrich us all.
• Outsourcing our thinking to AI will turn us into the Eloi free us to live fulfilling lives of some unspecified kind.

This, of course, is all hype and lies. The ‘AI’ systems we have almost certainly aren’t a path to intelligence of any kind. The most likely use they have is the use they’re being put to now: weaponized disinformation machines. Even without intelligence or any real possibility of reaching it they are busily destroying our ability to think.

But let’s assume the glorious AI future happens. Let’s assume that it’s not just a vast speculative bubble to make some plutocrats even more pluto, and that we will have access to real, cheap, machine intelligence in a few years. Let’s ignore the possibility that these intelligent machines decide to eliminate or enslave us, which is inconvenient to the politicians.

So: there are intelligent machines. The next question to ask is: are these machines conscious? This is something philosophers like to bullshit about at great length. But I am not a philosopher, so I’ll ask a simpler question: can you propose a test which will empirically tell whether these machine intelligences are conscious¹? If it helps, try to think of an empirical test which will tell you if other humans are conscious. Go ahead, I’ll wait.

You can’t, because there is no such test, for humans or machines. That being so, the only safe assumption to make is that yes, they are conscious. In other words the safe assumption is that they are people.

So, the glorious AI future is that we will all have intelligent machine assistants which we must assume are conscious and so are people. These people are bound to do our bidding.

There is a word for that: slavery. The glorious AI future is slavery.