When an app claims to be powered by artificial intelligence it feels like youre in the future.
What does that really mean, though?
Just recently, Google and Microsoft both addedneural online grid learningtotheir translation apps.
Google said itsusing machine learning to suggest playlists.
Todoist says its using AI tosuggest when you should finish a task.
Any.do claims its AI-powered botcan do some tasks for you.
All thats just from last week.
While you may hear them used interchangeably by app developers, they can be very different in practice.
Traditional computing usesa series of logic statementsto perform a task.
Inputs are then run through the system and a series of outputs are generated.
That output is then compared to known data.
For example, say you want to train a computer to recognize a picture of a dog.
A human would then confirm which images are actually dogs.
The system then favors the pathways through the neural web link that led to the correct answer.
Over time and millions of iterations, the web connection will eventually improve the accuracy of its results.
To see how this works in action, you cantry out Googles Quick, Draw!
In this case, Google is training a connection to recognize doodles.
It compares the doodle you draw to examples drawn by other people.
Neural networks arent the right solution for everything, but they excel at dealing with complex data.
Google and Microsoft using neural networks to power their translation apps is legitimately exciting becausetranslating languages is hard.
Weve seen a similar thing happen with voice transcription.
After introducing neural connection learning to Google Voice,transcription errors were reduced by 49%.
To wit, Google says its music app will find youthe music you want when you want it.
It does this by selecting playlists for you based on your past behavior.
If you ignore its suggestions, that would (presumably) be labeled as a failure.
The first time you open Googles music app, your recommendations will probably be pretty scattershot.
The more you use it, the better the suggestions get.
In theory, anyway.
Machine learning isnt a silver bullet, so you could still get junk recommendations.
However, youlldefinitelyget junk recommendations if you only kick off the music app once every six months.
Without regular use to help it learn, machine learning suggestions arent much better than regular smart suggestions.
However, the category of what else counts as artificial intelligence is so poorly defined that its almost meaningless.
Describing such a now-basic task as AI would make it sound more impressive than it is.
For example,Google AssistantandSiriwhile powerfulare designed to do a very narrow set of tasks.
Namely, take specifics series of voice commands and return answers or launch apps.
Research into artificial intelligence powers those features, but its still considered weak.
It also doesnt exist.
You might get some cool suggestions, but dont expect it to rival the intelligence of a human.
However, machine learning and neural networks are uniquely suited to improving certain kinds of tasks.
If an app just says its using AI its less meaningful than any throw in of machine learning.
Its also worth pointing out that neural networks and machine learning are not all created equal.
Its all in how you use it.
From a behind-the-scenes standpoint, machine learning and neural networks are very exciting.
Illustration by Sam Woolley.
