The Future’s Not as Good as It Used to Be

By Wil Forbis
July 1, 2018

Like every red blooded American, I’ve always had a healthy optimism about the future. Tomorrow holds the promise of not just change but change for the better. To verify this we need only compare today to the past. Aren't things much improved from 50, 100 or 500 years ago*? 

*I should state that, yes, I am aware of the various pontificators and academics who argue that past periods may have, in fact, been better than now. But, if only for the sake of the set up to this article, I maintain my optimism.

My anticipation for the future only increased when I discovered Reddit's Futurology section a few years ago. At that web forum one can find links to reports on all sorts of science fictiony technical advances that are occurring as we speak. Humanity is making great progress in health care, transportation, gene therapy and space travel, and, best of all, these advances should pay off in my lifetime. I only have to wait a few more years and the endless bounty of our utopian tomorrow will arrive.

Or so I then believed. But now I seem to have settled down from my futuristic fervor. Many of the wonders of the future that seemed so close at hand are stalling in their tracks. Consider the following...

Tesla Losing Speed
No company better captures optimism about the future than car manufacturer Tesla. The company, led by visionary Elon Musk, is aiming to upend the automobile industry by bringing electric cars to the masses. As a result the company is currently valued at 46 billion dollars.

The problem is that Tesla has repeatedly delivered late and/or underdelivered. Their initial product, the Model S, was announced in 2008 but arrived in 2012.  A later model, the Model 3, was promised in 2010 and finally arrived in 2017. (Many customers who placed deposits on the Model 3 are reportedly asking for their money back.) And all this is happening while Tesla is burning through a billion dollars a quarter.

Over the years Musk has been presented himself as a sort of technological sage who keenly sees a (mostly) bright future where others do not. But his latest behavior on Twitter has shown him to be more like a big man-baby who gets pissy when people start to question the narrative he's been selling.

I.B.M.'s Watson Disappoints
For years now we've all heard encouraging rumblings about computer intelligence and data mining. The idea, in broad strokes, is that computers will become better and better at solving problems and they will have bigger and bigger troughs of data to use for their analysis.

An area one would hope to be affected by these advances is medical diagnostics. If computers can take into consideration a full range of data about a patient (all test results, occupation, extensive medical history, etc.) and piece together a diagnosis with accuracy surpassing a human doctor, well, that would be a very good thing. Encouraging results have already been shown in A.I.’s diagnosis of skin cancer.

I.B.M. has long touted Watson---an A.I. technology which famously won the game Jeopardy in 2011---as a potential big player in the medical diagnosis world. The problem is, like Tesla, Watson ain't wowing 'em anymore. A recent contract with the M.D. Anderson Cancer Center was cancelled because, as Forbes notes:

...IBM fell into the trap of over-promising and under-delivering. “IBM claimed in 2013 that ‘a new era of computing has emerged’ and gave Forbes the impression that Watson ‘now tackles clinical trials’ and would be in use with patients in just a matter of months,” Freedman noted.

As to whether Watson will ever be useful in clinical situations? “This is hard,” opined Stephen Kraus, a partner at Bessemer Venture Partners. “It’s not happening today, and it might not be happening in five years. And it’s not going to replace doctors.”

When it comes to life saving computer diagnostics, I think most of us are anxious kids in the car: "Are we there yet?" The answer is, "Shut up and drink your soda!"

General Intelligence A.I. Far From Site
Watson is one example of computer scientists trying to develop smart machines. This development process has forced a reckoning with the question, "What is intelligence?" Rather than being one thing, intelligence is best thought of as a collection of skills. Artificial intelligence that encapsulates a specific skill is called narrow or weak A.I. One example of this would be the iPhone's Siri program that can answer basic questions like "When do the movies start?" but not more abstract ones like "Is there any point to this seemingly fruitless struggle we call life?"

Weak A.I. has its value but what scientists would really like to develop is strong A.I. This A.I, also known as Artificial General Intelligence (AGI) is, as Wikipedia explains, "the intelligence of a machine that could successfully perform any intellectual task that a human being can." AGI could make connections between data and calculations from a wide spectrum of knowledge and disciplines, for example, using information drawn from patterns found in sports statistics to make assertions about political trends.  

It would be great if AGI was just around the corner. Unfortunately, a number of headlines are appearing, like…

AI researchers are halting work on human-like machines

and...

Artificial General Intelligence Isn’t Around the Corner

It's not that researchers are giving up on AGI entirely, but the emerging consensus seems to be that the current methods of achieving it are falling flat. So it's back to the drawing board.

The Map of Life is Looking Murky
Anyone remember the Human Genome Project? In 2000 scientists Craig Ventner and Francis Collins stood with Bill Clinton in the White House and announced they had sequenced the DNA of the human body. The results of this effort were going to be monumental, as this Independent article written at the 10 year anniversary of the sequencing observed: 

…Bill Clinton said that the completion of the international Human Genome Project meant scientists were now learning the language in which God created life, while Tony Blair in Downing Street said that it represented a 21st-century revolution in medical science whose implications will far surpass the discovery of antibiotics in the 20th century. The human genome was going to change the way medicine is practised. It was going to reveal the hidden hand of cards dealt by our genes so we could cheat our genetic destiny.

So here we are now, almost 20 years past this announcement. And the results are… well, mixed. There have been some advances made, but I think anyone who fully appreciated the optimism in the air during 2000 would be disappointed. Genetic medicine still holds promises and those promises will eventually come through, but not necessarily soon.

This delay is partly because the whole mechanism by which genes pass on traits through generations was poorly understood before the genome sequencing (and perhaps still is.) For one thing, it turns out we have far fewer genes than we originally thought. We’ve also learned that gene expression (e.g. when genes “turn on”) can be affected by various mechanisms such as exposure to drugs/chemicals, diet and aging. (The discipline that studies these mechanisms is known as epigenetics and it vastly complicates the 20th century’s more simplistic view of how DNA works.) Additionally, before the genome sequencing, scientists had the view that a single gene simply provides the code to produce a single protein. In reality, some genes produce not proteins but short strands of RNA and what this RNA does is poorly understood.

All of this complexity has lead to a realization captured by cell biologist Mel Greaves. "We fooled ourselves into thinking the genome was going to be a transparent blueprint, but it's not."

In short: there’s a lot of work left to do. 

Crime Prediction Software Sucks
It may surprise you (it surprised me) to learn that courts are currently using software to predict the likelihood of certain people commiting crimes. This software---a form of AI called a risk assessment algorithm---is used for determining who gets bail or parole. The software accomplishes this by analyzing a defendant’s attributes such as age, gender and prior convictions and using that information to predict his or her potential for recidivism. 

It sounds like a smart idea, and one that could use a computer’s inherent impartiality to offset the biases of humans. Unfortunately, as Wired notes:

[Dartmouth College researchers Julia Dressel and Hany Farid] found that one popular risk-assessment algorithm, called Compas, predicts recidivism about as well as a random online poll of people who have no criminal justice training at all.

"There was essentially no difference between people responding to an online survey for a buck and this commercial software being used in the courts," says Farid, who teaches computer science at Dartmouth. "If this software is only as accurate as untrained people responding to an online survey, I think the courts should consider that when trying to decide how much weight to put on them in making decisions."

 

Facing the Limits of the Future
Now, we really shouldn't be surprised by all this. Science is hard. Creating technology that can have real, meaningful impact on people's lives is hard. And we have all seen the futuristic predictions from the past; farcical forecasts about how we would have flying cars and apartments on Mars by the year 2000. Of course that never happened and instead we got the steaming, feces-laden disaster zone we call the present. If we fairly assess how we misjudged the future (in the past) we can only come to one conclusion: the bright, techno-utopian tomorrow we've all been dreaming about is going to take a long time to get here.

Oh well. At least robot prostitutes are right on schedule.

What do you think? Leave your comments on the Guestbook!

Wil Forbis is a well known international playboy who lives a fast paced life attending chic parties, performing feats of derring-do and making love to the world's most beautiful women. Together with his partner, Scrotum-Boy, he is making the world safe for democracy. Email - acidlogic@hotmail.com

Visit Wil's web log, The Wil Forbis Blog, and receive complete enlightenment.