On AI and the Turing Test

Recently the Internet got all fired up–did a software bot pass/not pass the Turing Test:

The Turing Test, proposed by uber-genius Alan Turing in 1950, is: Can a software bot convince a human it’s also human, via text chat?

My view is that Turing Test has always been malformed, humans are too easy to trick, passing test says almost nothing about software. “But Marc, Alan Turing was the genius of all time, and you’re just some dude on Twitter. What the hell, man?” That’s a good point!

Turing said something else that I think is far more relevant, which he announced loudly in the executive cafeteria at Bell Labs in 1942: “I’m not interested in developing a powerful brain. All I’m after is just a mediocre brain, something like the president of AT&T.”

And I think that’s actually what happened, and it happened in the form of enterprise software: the code that runs businesses today. Yep, that’s right: Actual AI is SAP, Peoplesoft, Oracle, Salesforce, Workday, Netsuite, Great Plains, and a thousand others.

AI turned out to be the last thing anyone expected. Banal.

Credits: Conversations with @cdixon @balajis @bhorowitz

Source: Tweets – 1,2,3,4,5,6,7,8,9,10


Technology as Superpowers

New technologies can be thought of as giving people superpowers, superhuman abilities that humans did not have before. First, let’s think about historical technologies through this lens.

  • Man-made fire: superpower to heat and cook things.
  • Electric lighting = superpower to be active and productive even when the sun is down.
  • AC = superpower to act/produce when too hot.
  • Steam power and mechanical engines = superpower to exert physical effort in the world way beyond human or animal muscles.
  • Planes, trains, and automobiles = superpower to travel far faster than feet, animals, or wind could previously carry us.
  • Telegraph and telephone = superpower to communicate faster and cheaper than messages could be physically carried.
  • Threshing machines, combines, and tractors = superpower to grow far more food than we could by hand.

Of course there are destructive superpowers: guns and bombs enable killing far more people than we could by hand. For 500+ years, we’ve collectively been radically enhancing capabilities of ordinary humans through technology superpowers.

New Superpowers Just In The Last 10 Years

  • Google, Wikipedia = superpower to ask any question, get any answer, as many as you want, for free — from the world’s total knowledge base.
  • Facebook = superpower to always be connected with everyone important in your life, regardless of geography, all the time, for free.
  • eBay, Etsy, Alibaba = superpower to take goods you make or want to resell to a global market of buyers, with transparent and fair pricing.
  • GPS + Google Maps + smartphones = superpower to never be lost, always be able to find anything, and always know where your kids are.
  • Spotify, Beats = superpower to listen to the entirety of recorded music in human culture as much as you want, anytime you want.
  • Lyft, Uber, AirBNB, HotelTonight = superpower to be able to easily move around and then stay in places with transparency and safety
  • Skype, Slack, Asana = superpower to form into teams and collaborate on projects with people all over the world regardless of geography.
  • Github = superpower as a programmer to build new software with unprecedented ease and power, on top of a global base of existing code.
  • AWS = superpower as a programmer to access a global supercomputer with miraculous power on demand for mere dollars.

These are only a few examples. There are probably 100 more new tech superpowers just in the last 10 years of similar magnitude. I am firmly convinced many people are fundamentally underestimating the power and potential of these new superpowers in the years ahead. We are only at the very beginning of understanding what people all over the world are going to be able to create and build from here.

New Superpowers Coming In The Next 10 Years

This is a difficult list to make as there are hundreds if not thousands of candidate superpowers on deck that could hit it big within the next 10 years.

  • Crowdfunding = superpower to instantly raise money from global customers and investors for millions of new ideas, products, businesses.
  • Quantified self = superpower to understand one’s own body continuously in real time, optimize health & wellness with high precision.
  • 3D printing = superpower to design and instantly manufacture physical objects with zero inventory, supply chain, or transportation.
  • Combine modern bio, 3D printing, & computing–>prosthetics and exoskeletons; superpower: paralyzed to walk, disabled to abled, bilnd to see.
  • Bitcoin & blockchain = superpower to instantly & safely transact, do business, engage in commerce & trade with everyone in the world.
  • Virtual reality = superpower to experience other places and times, real and created, serious and fun; telepresence replaces transportation.
  • Augmented reality = superpower to know a million times more about the world around you and everything in it; step function increase in IQ.
  • Self-driving cars = superpower to dramatically reduce the number of cars on the road, turn parking lots into parks, while saving millions of lives.
  • Drones = superpower to easily function in the air just like we do on the ground; see from the sky, deliver in the air, and fly in virtual reality.

Implications Of The Growth Of Tech Superpowers

First, some tech superpowers help us as consumers, but many upgrade us as creators, builders, inventors, designers, artists: *producers*. While it is true that tech superpowers can replace the need for prior manual labor, it’s also true that they enhance our ability to produce.

Both sides of the equation are critical, otherwise one collapses into the Luddite fallacy and holds back progress that benefits everyone. This is also why I think we need an ongoing vigorous social safety net, to help people bridge gaps in their lives as tech leaps forward.

Second, tech superpowers at this point are being applied more or less equally to individuals, governments, and businesses. In Orwell’s vision, government has all the tech superpowers. In our world, everyone seems to be getting them all at the same time. The same tech advances that enable NSA surveillance also enabled Edward Snowden to walk out of NSA with 1.7 million classified documents. The ability to use a web crawler and a thumb drive would have been the fantasy of all time for Kim Philby. Times have changed.

The impact of tech superpower upgrading of individuals, businesses, and governments radically shifts power balances between the three. The politics of the next 30 years will be in large part defined by the shifting impact of tech superpowers across individuals, businesses, and government.

Prior conflicts like Snowden, SOPA, the Arab Spring, and cybercrime are only open gambits — the real drama is yet to come.

Source Tweets Part 1: 1,2,3,4,5,6,7,8,9,10

Source Tweets Part 2: 1,2,3,4,5,6,7,8,9,10,11,12

Source Tweets Part 3: 1,2,3,4,5,6,7,8,9,10

Source Tweets Part 4: 1,2,3,4,5,6,7,8,9,10,11,12

A Coke is a Coke

For the following, substitute rapidly rising tide of products and services for “Coke”:

What’s great about this country is that America started the tradition where the richest consumers buy essentially the same things as the poorest. You can be watching TV and see Coca-Cola, and you know that the President drinks Coke, Liz Taylor drinks Coke, and just think, you can drink Coke, too. A Coke is a Coke and no amount of money can get you a better Coke than the one the bum on the corner is drinking. All the Cokes are the same and all the Cokes are good. Liz Taylor knows it, the President knows it, the bum knows it, and you know it.

Andy Warhol on Coca Cola

Source: Tweets – 1,2,3,4,5,6


Cycle Time Compression: The Most Underestimated Force in Determining Winners and Losers in Tech

Cycle time compression may be the most underestimated force in determining winners and losers in tech. The first clear instance of cycle time compression: Cloud/SAAS vs on-premise enterprise software. Cloud/SAAS development cycles can be far faster than on-premise software; single instance deployed instantly to all customers. Further, customers can try and adopt cloud/SAAS far faster than they can try and adopt on-premise software.

Implication: Cloud/SAAS is probably impossible to compete with for on-premise software across multiple product cycles.

The second clear instance of cycle time compression: Product improvement and customer upgrade cycles for phones vs TVs and cars. Consumers can upgrade their phones every 1-2 years, vs TVs at 5-8 years? Cars at 10-12 years? With phones improving by leaps & bounds.

Implication: At given point in time, your TV can be 4-6 years behind your phone; your car can be 9-10 years behind your phone.

Implication: TVs and cars will become accessories for phones, not the other way around. And it’s already happening: Airplay, Chromecast.

Interesting note: Web cycle times still much faster than mobile app cycle times due to restrictive mobile app store policies.

Source: Tweets – 1,2,3,4,5,6,7,8,9,10


Why Robots Won’t Eat All the Jobs

This is probably a good time to say that I don’t believe “robots will eat all the jobs.” The preceding tweetstream was to extrapolate the idea out all the way, not to make the case that it’s what’s going to happen.

First, robots and AI are not nearly as powerful and sophisticated as I think people are starting to fear. Really. With my VC/tech hat on I wish they were, but they’re not. There are enormous gaps between what we want them to do and what they can do. So there is still an enormous gap between what many people do in jobs today and what robots and AI can replace, and will be for decades.

Second, even when robots and AI are far more powerful, there will still be many things that people can do that robots and AI can’t. Creativity, innovation, exploration, art, science, entertainment, caring for others… we have no idea how to make machines do these.

Third, when automation is abundant and cheap, human experiences become rare and valuable. It flows from our nature as human beings. Examples: Price of recorded music goes to zero; live touring business explodes. Price of drip coffee drops; handmade gourmet coffee grows. You see this effect throughout luxury goods markets, e.g. handmade high-end clothes. This will extend out to far more consumers in future.

Fourth, just as most of us today have jobs that weren’t even invented 100 years ago, the same will be true 100 years from now. We have no idea what the fields/industries/businesses/jobs of the future will be; we just know we will create an enormous # of them. If robots/AI replace people for many of the things we do today, the new fields we create will build on a huge number of people then available. People 50, 100, 150, 200 years ago would marvel at the jobs that exist today; the same will be true 50-100-150-200 years from now. To argue huge numbers of people will be available but we will find nothing for them (us) to do is to dramatically short human creativity.

Source: Tweets – 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15


Human Potential

The flip side of the “robots eat all the jobs” theory not being discussed: The current revolution in the “means of production” going to everyone in the form of the smartphone (and tablet and PC) + mobile broadband + the Internet: Will be in almost everyone’s hands by 2020. Then everyone gets access to unlimited information, communication, education, access to markets, participate in global market economy.

This is not a world we have ever lived in: Historically most people in most places cut off from these things, usually to a high degree. It is hard to believe that the result will not be a widespread global unleashing of creativity, productivity, and human potential. It is hard to believe that people will get these capabilities and then come up with absolutely nothing useful to do with them.

And yet that is the subtext to the “this time is different” argument that there won’t be new ideas, fields, industries, businesses, jobs. In arguing this with an economist friend, the response was “But most people are like horses; they have only their manual labor to offer.” I don’t believe that, and I don’t want to live in a world in which that’s the case. I think people everywhere have far more potential.

Source: Tweets – 1,2,3,4,5,6,7,8,9


Consider the Tractor

Consider the development of the tractor, which mechanized virtually all of agriculture over the 20th century. Somehow new desires and demands sprung up for new kinds of manufactured goods, many of pure entertainment value, and people stayed employed and real wages kept rising.

Capital Eats the World – by Suresh Naidu

Source: Tweets – 1,2,3