As with any new area of technology, business understanding of AI lags the work being done by AI practitioners.  With the massive amounts of investments being made on AI, it is important, especially for executive decision makers, to look behind the curtain and get a better understanding of the constraints and applicability of AI applications to their business.

Three applications that have got a lot of press in the last year are AlphaGo Zero, Amazon Go, and autonomous cars. This post looks at the relative complexity, constraints, and cost to implement each solution, as well as their potential for disruption. It’s no surprise that AlphaGo Zero, which has shown the sharpest of results and is the least expensive to implement, has the narrowest applicability of the three. On the other end of the spectrum, autonomous cars have the potential to be fundamentally disruptive. They also represent the most complexity (what is the definition of “safe” in bits and bytes?) and are the most expensive of the three applications. While the technology is already having an impact on our lives, fully autonomous cars are still many years away.

A deeper examination of each, nonetheless, can provide insights on how to evaluate the constraints by which AI operates, and how it can have an impact on achieving business goals.

Uber – 13 miles per driver intervention

https://www.nytimes.com/2018/03/23/technology/uber-self-driving-cars-arizona.html

While Arena claims to reduce median turnover by 38%

https://www.economist.com/news/special-report/21739433-ai-changing-way-firms-screen-hire-and-manage-their-talent-managing-human-resources?frsc=dg%7Ce

1. Marketing and sales – $1.4 trillion
2. Supply chain – $1.3 trillion

https://www.economist.com/news/special-report/21739428-ai-making-companies-swifter-cleverer-and-leaner-how-ai-spreading-throughout?frsc=dg%7Ce

1. A black man born to parents at the median income would expect to end up lower on the income ladder than his parents, while a white man born to parents of median income would not. This is important, because it abstracts from dependency on how well-off one’s parents are.
2. Among children with parents earning at the 25th percentile, black boys had lower incomes in adulthood than white boys in 99% of neighborhoods

       https://www.economist.com/blogs/democracyinamerica/2018/04/broken-ladder

50 – number of driverless car companies in CA

0 – number of applications

No driverless cars in California? Why April 2 won’t make history

The focus on Federal taxes as a Robin Hood-type redistribution tends to focus on economic classes – is the middle class a winner or a loser, etc. Except, Representatives don’t represent classes. They represent states.  And, looking at the redistributive effects of taxes on a state by state basis, the Republican party comes across as just a machine in directing benefits their way.

If we look at net Federal spend (spend less taxes) by state, Republican states get a net benefit of $284B, while Democratic states are net losers by $216B.  The top 5 winning states include one Democratic state – Maryland – and 4 Republican states.  This reverses for the bottom 5 states, with only one Republican state – Ohio – represented, along with 4 Democratic states. Closer to home, California weighs in dismally short at $29B.  It will get worse as some of the relatively more punitive measures in the new tax legislation hits residents in Democratic states the hardest.  Specifically, on being able to deduct state taxes on Federal tax returns.

Maybe it’s no surprise Republican party colors are red.  Democratic states need better representation

The skinny repeal was to do away with the individual mandate where individuals are required to buy health insurance or face penalties. It is a bag of lemons.

George Akerlov won the Nobel Prize in Economics for a paper he published in 1970 called “The Market for Lemons: Quality Uncertainty and the Market Mechanism”. In it he said that markets could collapse if one side of the transaction had more information on the quality of the product than the other. For example, if a buyer could not tell apart a bad car (a “lemon”) from a good car (a “peach”), they would not be willing to pay a price greater than the average quality of the car they expected to buy. That would lead sellers of peaches to withdraw from the market, since the price they would get for their car would be less than the value they knew it to hold. And, we would be left with a market in which only lemons were sold.

This is obviously an example taken to an extreme, but does indicate the directional impact of doing away with the individual mandate. People who are more at risk would be more likely to buy insurance, leading to a riskier pool, and put upward pressure over time on premiums. Rising premiums would, in turn, push people who are on the fence, i.e., the relatively lower risk participants, to not seek insurance in the future. Thereby putting additional pressure on premiums. And so on.  Potentially leading to a non-functioning market.

Obamacare also does not allow premiums to be based on pre-existing conditions. This, in some ways, is the other side of the glove of the individual mandate.  In that it allows premiums to be priced purely on age, and not on other health considerations. We could argue the merit of that. But then, we abut against philosophical considerations on insurance. The purpose of insurance is to pool risk. I may not be sure on an outcome to me any given year, but by putting myself in a much larger pool, I tie myself to the security of the group outcome. Obamacare helps insure that.  Pricing based on individual health considerations, however, reduces this pooling of risk. And taken to the extreme, each of us gets a health plan tailored to our specific needs, and pays the premium associated with it. Which really means no insurance at all, at least from a pooling perspective. Let the chips fall where they may… And given the cost of health care, very few people really want that.

Which brings us full circle to the fundamental problem faced in the country with health care. Costs have been on a tear pretty much since the 80s. The skinny repeal does not address this increase in cost. In fairness, Obamacare does not either. But, neither can it be blamed for it. Cost is a pre-existing consideration that is being ignored by both our great parties.

A pedestrian was killed in Tempe, Arizona by an Uber autonomous car.  In 2015, Governor Doug Ducey enticed the self-driving car industry to Arizona by executive order clearing the way for testing in the state.  Last month, he updated this order touting Arizona’s “business friendly and low regulatory environment”.  Following the crash, Uber has stopped all real-word testing of its autonomous cars, which were happening in San Francisco, Phoenix, Pittsburg and Toronto.  The accident is now in the crosshairs of both the U.S. National Highway Traffic Safety Administration and the National Transportation Safety Board.

The recent Cambridge Analytica revelations on Facebook data to help Donald Trump’s campaign is ill-timed for autonomous car companies.  And is forcing regulators to increase scrutiny on the level of self-policing that has so far been granted to tech companies generally. The fatality and recent privacy breach revelations will almost certainly adversely impact the pace of autonomous car technology advancement in the U.S.

There are at least two broad black box areas regulators will want to examine and ultimately address.  One is conceptually straightforward, while being technically bedeviling.  Autonomous cars are trained using AI methods such as deep learning on massive amounts of data to interpret and react to driving conditions.  However, unlike traditional statistical predictive methods such as regression analysis, deep learning does not easily lend itself to transparency of decision making, which leaves it with an air of magic about it.  This reality will make it difficult for regulators to communicate with an increasingly skeptical public.

The other issue is philosophically much more challenging.  Inevitably, autonomous cars are going to be in situations requiring them to make an instantaneous choice between a set of bad outcomes.  For example, the decision when an autonomous car is faced with a choice of modest damage to itself versus more material damage to its surroundings.  Even more fundamentally, what happens when lives are at stake?  How will the car measure tradeoffs and react to them?  At some level the processes for making these ethical decisions must be programmed into the car.  The public at large is unlikely to accept a Google, BMW, Ford, or an Uber unilaterally making such decisions.  Recent headlines on Cambridge Analytica that erode public trust in tech companies, and, now, a self-driving car fatality will force a bright spotlight at the core of autonomous vehicle systems.

The rule of law and consumer protection is a strength of the U.S.  At the same time, these strengths could prove to be an impediment in the race for global leadership in the development of autonomous cars, and AI more broadly.

Hype, hope, and modern science fiction cinema have captured our imaginations and created significant expectations around the deployment of machine learning and AI in commercially available products in our near future (e.g. self driving cars). However, don’t expect to be transported to Machine City in Matrix Revolutions (very cool, in its own way) any time soon. The rise of the machines will have to wait. Expect to see plenty more incremental steps such as collision avoidance and semi-autonomous driving (hopefully, with limited significant human cost as we adjust to the reality of the state of the technology today… see – tesla on autopilot slams into parked firetruck), and a few eye catching advances such as Amazon Go (how cool is that!?).

At least some of the public attention on deep learning is from the name, aided by some great science fiction filmmaking and computer graphics, evoking images of an inward looking, thoughtful, mathematically robust, and far less bloody, Ex Machina thing.

The underlying concept of deep learning, however, is not quite that deep. In fact, it is conceptually very similar to simple regression analysis, which evaluates and describes a relationship between variables – say between gender wage gap and Donald Trump’s popularity (true story).

Sources: National Women’s Law Center, Gallup

The typical way to try and describe this type of relationship is to fit a straight line on the data that best summarizes the relationship. Why often a straight line? Well, computationally a straight-line relationship is easiest to try and fit. Similar ease of computation considerations are true for deep learning models as well.

A major consideration is to decide what line best fits the data. We choose a line from options that minimizes an aggregate measure of difference between the actual data to that predicted by the line – the line of best fit. That is the essence of deep learning – computational techniques, albeit far more complicated, that try and establish relationships between variables by minimizing some measure of error.

A key difference is that deep learning tests out relationships using additional layers of variables. Hence the term “deep”. More layers = deep. We do not make any assumption on what these variables are, just how many layers, the number of variables in each layer, and the mathematical relationship between the layers (parallel to the example of a straight-line relationship above). There is no magic to these choices. You select a configuration based on what is computationally feasible and provides the best results, and then you have deep learning. Still some ways away from Machine City.