Artificial Intelligence in the News

I suppose we should get used to seeing topics like this discussed in the mainstream news. For the record I consider myself pretty well informed on A.I. technology and have tried to keep up with developments since, well, the late 1980s. I have also done some similar circuit design work. I would probably also be considered an A.I. “skeptic” although I consider myself more of a “realist”.

In this story, some Google engineers have some issues with recent claims of using A.I. to build circuits. I’m sure this is a deep, inside baseball argument going on here, but not to be allow to be discuss this publicly is a bit strange. Again, for the record, I think there are many overblown claims for A.I. (self driving cars and IBM Watson, I’m looking at you) but maybe the real problem is that the overblown claims get all the attention.

Nuclear Power and Water Use

Another issue with using nuclear power, especially in the western US. I originally saw this in an article about an Arizona nuclear power plant which is behind a paywall. A Google search turns up some numbers. From a UK article.

“A large nuclear power plant may use up to 1 billion gallons of water a day and, for this reason, they are often built next to rivers, lakes or oceans to utilise the bodies of water.”

Nuclear power and its water consumption secrets

Elon Musk and Twitter

Lots being written about this, most of it missing the point. In the short term, a single, very wealthy person controls one of the most popular social media platforms in the world. What can you do with this sort of platform? In the short term, promote your brand(s). Get people to buy more Tesla’s. But perhaps he has political ambitions. Can Twitter sway any election, anywhere in the world? Probably. So imagine your favorite (of least favorite) politician owning Twitter. Afraid yet?

And this talk of opening up the algorithm is a distraction. What I would want to see is the data. Who is getting what postings, and who is paying for special access (advertisements pretending to be postings).

What can be done? I dunno. Someone built a bomb, now you have to defuse it. Get rid of all anonymous and robot accounts? Flag paid content as advertisements? Apply laws for libel, slander, terroristic threats, etc. to postings. Basically arrest people who make threats to other people on line, as would happen in real life, on the phone or through the mails. Provide data on content accuracy. Flat earth people won’t get their message out, but is that really a problem? It will cut into Twitter traffic and revenues, though. Lastly, treat it like we have treated other public communication infrastructures, i.e. telephones, television, and regulate it. Make them do some or all of the things above.