Apparently this algorithm can be used to modify and corrupt the machine learning processes of artificial intelligence. Big companies are already using AI for a variety of things, so a corruption could hurt a lot of people and cost a lot of money.
But this is not actually the biggest problem with AI. Turns out that if the programmer is racist or biases, the AI will be trained to be racist or biased.
And if the owner of the AI happens to be an intrusive asshole, like Facebook or other social networks, then guess what? The AI will be intrusive, assshole-ish and will be ten times worse than people can ever be.
When it comes to AI, which is not yet in existence, no one programmer writes the code. In fact, when it comes to any complex software system, such as Windows, no one person writes that code either. I was a Microsoft Developer for years, and worked closely with their engineers and programmers. I found out the hard way that every problem I had required them to reach out to the one programmer that worked on that particular part of the software. Besides, I’ve looked at more software programming than most, and have yet to see any kind of social bias built into any of it. If someone has actual code that is anti-, or anti-, why has it never been posted in a public media site?
In this case, we are talking about neural networks and training, which to the layman is synonymous with AI.
And yes, it is not the software that has the bias it is the training data entered. At an AI conference a few years ago they described what they called “racist” software for predicting recidivism in crime in an area. They included race as a factor, which made the results racist. It predicted that a particularly nasty white guy would be less likely to commit another crime vs a less nasty black guy, because it grouped the white guy with the more affluent white neighborhood next door. Obvious, in hindsight.
thats utter nonsense. including race as a factor in an algorithm is no more racist than including residential addresses or average incomes in the same data sets.
there are actual disparities between the races when the commission of crime is studied.
in fact, if you parse a database with regard to the nature of crimes committed and then organize those by the racial background of the individual responsible, you will find that certain crimes are perperated in many cases more often by one group more than the other.
similiar results are found if rather than race the results are filtered by cultural background or geographic location. additonally, these results will very often coorelate with the results found by filtering by racial factors…
thats the real…