AI – Are we there yet?

One of the things that an Internet of ‘things’ is going to generate is a huge amount of data.
If every home in our suburb’s thermostat was somehow linked to the Internet and a single value, once every 15 minutes was stored in a database, you would have a few things. You would have a lot of data. 1440 data points a day per house for starters….
Data storage is cheap, so lets forget that aspect.

Ok, so you have the data. Now what?

This is the challenge facing IoT. Making sense of large data sets.
Not just making ‘sense’ (what exactly does that mean to a computer?), but trolling through the data and coming up with something that will improve the quality of life of someone or something as a result.

In English?

Ok. That sort of data (and I just made the example up on the fly) can provide all sorts of helpful insight to weather (people feel cold, so they turn the heat up – oh, wait, we are in SoCal, they are ‘never’ cold) for example. So you would be able to tell the mood of the place.
Power. By looking at all the data over time, you could start to easily tell when the bulk of the suburb comes home and turns on the AC, thus putting a load on the electrical grid supplying that area.
Both those are reactive…..

What if you looked at the data, for say, a year. Or two.
Just had a computer sit there with some clever math and looked at trends and stored those trends and matched the daily and weekly trends. That same computer ran checks on itself, and made sure that the trends it saw were repeatable.
Now, with some certainty, the computer can now predict.

AI. Artificial Intelligence.

Pie in the sky?
Not so much – you probably use it every single day. Many times a day.
Google search.

Every time you click on a link that you think matches best the search term you enter into Google (or Bing, or Yahoo or DuckDuckGo), you reward the AI that Google has built.
Billions of times a day, the math formula is rewarded or punished.
It learns.
If offers better search results.
The humans are happy.

Now read this;

I have more to say on this topic at another time.