I find the buzzword business quite frustrating sometimes. It seems to enjoy messing with meanings and having me gasping over its grammar. At first glance, artificial intelligence is a contradiction in terms. Intelligence is, after all a human ability to acquire, process, and apply knowledge and skills to situations, and so move the frontiers of knowledge outward.
If it were artificial, surely we would still be sitting up in trees and eating chestnuts.
Intelligent Answers for Things that Hold Us Back
I for one am not in a tree eating chestnuts. Although René Descartes frustrated the hell out of me at college, after I rashly signed up for philosophy to earn enough credits to complete the course. We saw the ancient head of department only once. I never did find out whether he was still alive when I dropped out. The venerable prof said, “Welcome to philosophy. If you are looking for answers, you won’t find them here. Answers have no place in this class, only questions.”
In retrospect, they should have included Data Science as a Foreign Language in the curriculum; apparently we need to start learning the jargon, as adised by Kirk Borne.
For our money, he made us study Descartes’ scribblings. René was a strange fellow because he ‘invented philosophy’, although Socrates and Plato dabbled in it two millenia earlier. Oddly, he had an answer to his conundrum whether he existed. He exclaimed “cogito ergo sum, je pense, donc je suis; I think, therefore I am” and concluded this opened the possibility of the creator existing too.
Machine Learning Versus Artificial Intelligence
Forbes’ R L Adams says machines have not yet evolved to the point of being capable of original thought. Computers do however get results sooner from big data. They have crept up to us in ways we may not even notice. Some are incredibly annoying. Like the spell check on Word that converts my typos into flights of fancy, and makes me want to either laugh or cry. Artificial intelligence bah humbug I say. We have a way to go before we crack champagne on that bot’s head.
Some Less and More Useful Examples of Machine Learning
‘AI’ is customer relations management by computer. Its machines track our thoughts and preferences as evidenced in our behavior, and make calls to action. Like any good sales rep, this generally happens within their paradigm. It’s often more a case of ‘will you come into my parlor said the spider to the fly.’
- The Google search engine has a habit of showing us results it thinks we ought to see. This is all very aboveboard they told us, and then they introduced paid advertising. I am not convinced this thought policing is a good idea at all. The Chinese seem to have their minds around internet censorship already. Now the American Federal Communications Commission wants big companies – and potentially government departments – to bribe their way into preferential line speeds.
- The spam filter in my mailbox is a regular source of irritation too. It randomly assumes a mass mailing is spam, and dumps it somewhere I have to remember to look. This is unintelligence at its worst. Only bloody-minded machine learning would fail to realize my medical insurance has a million members. Thus it would be mass insanity to send a million separate emails asking for nominations to the new board.
Despite this rant, I have to admit there are a number of useful machine-taught bots out there.
- I confess having a soft spot for Alexa. She listens to me from anywhere in the room, and helps me find what I am looking for among a vast crowd of opportunities. I am using the home automation module to manage my HVAC, and monitor my refrigerator temperature too. (I for one am anxiously waiting for her to vacuum the house, note from the editor)
- Cogito is another personal assistant I rather fancy. Cogito helps telephone professionals deliver ‘superior service’ by detecting human signals and providing live behavioral guidance. Expert System says it bases its cognitive capabilities on “intelligence algorithms that mimic the human ability to think at the speed of current technologies“. Cogito means ‘je pense, I think’. We have Descartes to thank for that burst of intelligence.
- Of course, we also have social media ‘to thank’ for extending our range of contacts. With Facebook, the level of ‘intelligence’ seems limited to suggesting friends of friends (or so it feels). LinkedIn uses more advanced algorithms that are more often than not off target. I live in a small village called Umtentweni nobody knows of. When LinkedIn offers to introduce me to kindred spirits down the road, it’s a miracle if we are in the same country.
- That said, I adore using Google Maps on Android. Provided someone reports them, it can analyze my planned route, tell me about road construction projects and traffic accidents, and suggest faster alternatives I may not have thought of. This comes closer to the form of artificial intelligence I want. That’s really useful information I can use, available online, on time.
What Could Be Around the Corner with AI
I could rabbit on longer, but you get the drift. We don’t have all of it yet, but we do have machines with fast learning and computing ability. We enter another level when we think of unleashing the beast to make automated decisions on our behalf. It would be bad enough if Amazon automatically dispatched the auto spare it thought I wanted, and billed me for the privilege.
However having a machine decide when to trump the North Koreans for having gone a step too far gets scary. A bit like getting into a pilotless air taxi and hoping there is enough electricity in the batteries. Please don’t get me wrong. I am not against thinking machines .I would agree with Ms Maria Johnsen that fear of AI is counterproductive.
In my opinion #ArtificialIntelligence #AI is misunderstood.
Creating fear from A.I is unhealthy for #technologial advancement#MariaJohnsenpic.twitter.com/kNgWVMAhcs— Maria Johnsen (@iMariaJohnsen) September 6, 2017
In fact, thinking machines that cough up enough actionable data to make our marketing jobs easier is a form of inbound intelligence we all can (and should) embrace.
My concern is we already have some artificial intelligence we did not ask for, and I doubt our collective ability to manage it. Forbes is one of many to predict AI robots will take over our jobs. We already replaced bank tellers with auto tellers although the queues still seem as slow as ever. It could be a great thing if some of Jean-Marc Côté’s predictionscame true from a century ago. We just need to decide what to do with all those people we replace.