fbpx
Stop feeding the machine

Careful with that A.I. Eugene !  /  How NOT to A.I.  /  Stop feeding the machine.

Stop feeding the machine. You may have heard a lot about A.I. recently, as lots of money is being thrown at large A.I. projects to explore the space and open new possibilities with computing. While at the moment it seems like a solution looking for a problem, in time it could prove to have great potential.

There are many types of A.I. models, which can be general or specific to tasks, such as language models and audio-visual models.

There are three things every A.I. usually requires to function – aside from money –

Huge computing power

Lots of chips and lots of memory, the more the better. The more factors an A.I. can consider the better the results will likely be.

Huge amounts of information

Every A.I. will have a large bank of information from which it draws conclusions and makes decisions or predictions.

People, you, me and everyone

Without the people, there would be no data or content, and no reason for the A.I.

The A.I. is unthinking and uncaring

The A.I. does not understand, and it does not care if it is wrong. The output provided may well be wrong or complete nonsense. You should proofread anything it makes, even if you used the A.I. to proofread it. Also, consider what information you feed it, are you sharing personal information or company data? Do you know how that will be used later? Probably you don’t. Do not just fire sensitive information into the giant machine, who knows where it goes.

It should go without saying, but you should not use A.I. for nefarious purposes, create misleading or harmful content. However, by the same token, you should be increasingly on guard against such content as there will be more and more people trying to do just that. Criminals can now automate their scams and phishing attempts on large scales and in any language. Pictures and videos can be faked, fake humans can pass the Turing test pretty easily and are able to generate fake identities and converse with you.

Hopefully the technology will be used for great purposes, maybe a universal translator, a breakthrough in science or maths too complex for mortal brains or a superb medical diagnostician with all the medical knowledge. Whatever happens, it isn’t going away.

Picture by Dall-E2

Ben Parker – byline and other articles