AI: past, present and future on Strategy
This past wednesday, I attended a Tesla workshop at ISCTE business school, on green solutions. The day before I also participated in the “Tesla case study: bring Gigafactory to Portugal” event at ISCAD administration school. Friday evening I joined the discussion panel at Startup Sintra Meetup on IoT vs Industry, with a live streaming on youtube. All of the above implicated large doses of reading thru subjects for myself on a whole lot of subjects. Tired already? BTW last week everybody was discussing Teslas’ 4Q and Full Year 2016 update on shareholders, and CFOs’ shorter-than-expected leave, another time consumer (link below).
On Feb’17th Quartz released an interview with Bill Gates, in which he said on “robot taxes”: they should be raised in order to balance the previous asset that was also being taxed (us humans!) and maybe we should consider deploying robots in substitution for human activities at a lower rate than previewed, due to the social impact of such wide scale shift in a near future. Dare to say Mr. Gates maybe right on the later. The former will probably be easily solved by a card dealer of sorts.
On that subject comes to mind that it was in May 11, 1997, that Mr. Garry Kasparov was (clearly) defeated by IBMs’ Deep Blue. Thats right, nearly 20 years have passed on this historic event, when mans’ most highly ranked chess player was tilted by a “machine”. Not a regular machine by 90s’ standards I will add.
By that time, Artificial Intelligence was still scientific jargon, or a lost Asimov chapter from some decades before. By todays standards, the term “artificial intelligence” is applied when a machine is able to mimic “cognitive” functions that humans associate with other humans, such as “learning” and “problem solving” (known as Machine Learning).
Machine learning is the subfield of computer science that gives computers the ability to learn without being explicitly programmed (Arthur Samuel, 1959). Machine learning is employed in a range of computing tasks where designing and programming explicit algorithms is infeasible. Machine learning has around 15 different approaches or models (for now). Some of them have immediate (constructive) results for users. Say, deep learning can predict what the higher-resolution photo might look like, and add missing details on a low-res pic — great!
Other fantastic examples I saw last year, on Web Summit 2016 in Lisbon, were presented by Autodesk CEO Carl Bass, on a off-road buggy frame optimisation, a heat exchanger and a simple chair. Final studies provided a bio-like geometry for all subjects - really exciting and definitely a game changer for product design and product optimisation.
More and more examples add up as you search the web, and some will probably have a more deep impact on society that the ones I mentioned.
How to use that(those) powerful tool(s)? I encourage you to take a look at Andreas Markdalen post on “advanced design” where he proposes a collaborative model approach to Design (link below), as an example.
On the automotive scope, Tesla Motors jumps as a regular #01 reference on the subject of self driving cars. But they are not alone. As of 2016, there are over 30 companies utilising AI and ML into the creation of driverless cars. Level 5 SAE automated driving will be probably attained around 2028 or so according to latest studies (not to be confused with marketing catch phrases from OEMs that typically announce some-kind-of-assisted-driving much earlier).
So, we are now faced with the fact that machines are smarter, faster, cheaper and more reliable than humans. During the last decades humans have been arguing on the “what will be left for us?” subject. It started as “cognitive human only” label, and as we stand now, people speak about an “empathy only humans have” label. For how long?
What about the next status quo? What will be left for humans? Just to worry about “soul or spiritual issues”, maybe?
Between today and that future stage — whatever it is — we will still need to worry on manoeuvring our projects and our companies. Not sure when that will change, but you need strategy on your side, as always. Here is what Rober Grant says about the chess example:
“Consider the 1997 “man-versus-computer” chess epic in which Garry Kasparov was defeated by IBM’s “Deep Blue.” Deep Blue did not need strategy. Its phenomenal memory and computing power allowed it to identify its optimal moves based on a huge decision tree. Kasparov — although the world’s greatest chess player — was subject to bounded rationality: his decision analysis was subject to the cognitive limitations that constrain all human beings. For chess players, a strategy offers guidelines and decision criteria that assist positioning and help create opportunities.”
Mr. Kasparov did loose the 1997 battle, but surely his historic feats and his world position were attained using strategy by all means. Even when times are changing, strategy is always on your side, think about it. Aviva, one of the worlds biggest insurance companies recently poled their 16.000+ employees to ask them if they thought their role could be filled by a robot. In case the individual say “yes” he or she will be prompted to start training for alternative positions — you have to admire that. Aviva knows that “resources and capabilities” are one of the company pillars for a successive strategy.
The future is built on todays’ decisions — thats for sure.
Share, comment and like. Enjoy.
References:
Andreas Markdalen, The 44th Move — Deep Blue, Kasparov and the Future of (Visual) Design,
Higher-resolution photos, https://github.com/alexjc/neural-enhance
Grant, Robert M., Contemporary strategy analysis/Robert M. Grant.
Teslas’ 4Q and Full Year 2016 update
Deep Blue versus Garry Kasparov
