Perspectives: Opinions from our network of advisors, investors, operators and analysts on the risks and opportunities they see.
Frank Barbieri, an investor and co-founder of the San Francisco startup consultancy Ecolese, has worked in senior positions at Microsoft, InfoSpace and other technology firms and guided several to successful exits, including YuMe’s IPO in 2013.
Barbieri’s latest role is as Chief Development Officer at PebblePost, a New York-based Artificial Intelligence (AI) and data-informed direct mail marketing firm. But Barbieri’s network in Silicon Valley includes many leading developers, and he has watched with concern as AI, the driver of so many of today’s new models, has been cast in doomsday terms by everyone from Hollywood scriptwriters to economists to labor union activists.
He offered this dose of optimism to Karma Network’s Contributing Editor, Michael Moran.
Michael Moran: Frank, it seems like only yesterday that AI was on the tip of everyone’s tongue, a kind of tech industry palliative for everything that ails humanity. Now AI seems as often to feature as a job destroying, killer robot-enabling threat to human survival. Is AI getting a bad rap?
Frank Barbieri: This is something I’ve been thinking about a lot. A couple of our investments at ecolese, my consultancy, are aligned around, is the future of work, and how automation and artificial intelligence may impact those things.
There’s a real fear amongst the undereducated and unskilled labor force that they will lose their jobs, which is a reasonable and valid response, but I think it’s also possible to imagine a future in which, after removing some of the constraints that exist because of a manual labor challenge, you can imagine new business models emerge.
People think that when Uber goes fully autonomous or when there’s autonomous trucking, two to three million jobs will disappear from the driver pool across the United States. But I think it’s harder to calculate what the release on the barriers of transportation might mean for new economic models.
Michael Moran: I think many people find it easy to worry that robots and AI generally will displace humans in the workforce. What is the mitigating factor they’re not seeing?
Barbieri: When you have frictionless logistics crisscrossing nationwide, and moving goods from New York to San Francisco can be done at twice the speed for half the cost, that might engender new business models. You can imagine a flourishing of micro-manufacturing across the country, because it’s easier for individuals to service a larger market at a competitive price.
One possible future scenario that I think is absolutely valid and could actually be a boon for the economy over time, is that humans — particularly the unskilled labor force — shift their employment from giant hubs to much smaller and community-based businesses.
So while I don’t want to be Pollyannaish about how technology is deployed and what it accomplishes, it’s really never been proven that the rise of technology has been a net negative for humanity.
I think that at the end of the day, removing barriers from human potential is a great thing for commerce. It’s great for business, and it will be a great thing for humanity.
Michael Moran: How is AI going to affect inequality among groups of people? Is it going to help the problem, or just intensify it?
Barbieri: When thinking about AI, I always try to imagine a world in which there are fewer boundaries. I also think that that thesis presumes “ownership” of an intelligence, so you will see large hubs of powerful intelligence that have an ownership component.
We’ve seen throughout history that the most stable and effective commerce model for any resource is in a marketplace. So I think that it can be a great leveling ground if serious intelligence is available to the average person to help solve a problem, to plan or to compete with another AI [that may be performing a job in an unsatisfactory manner]. And you can have a periodic or a rental model of the intelligence, just as you do with IBM’s Watson computer, but you’ll have a far broader market with a lot of competition. I think everyone will be able to participate in that.
So does inequality still exist? Could it accelerate inequality? Absolutely. But I think access to things that were never available before will also ride alongside that, so there won’t be just a pure “walled garden” ownership of these new intelligences, but, instead,a lot of folks who get access to them. And the effects of that part, I believe, are unpredictable.
The future of customer service is one AI negotiating with another AI, where the seller AI is trying to emulate empathy while still maintaining profit, and the buyer AI is trying to exploit that empathy.
You can imagine a world in which you had a bad transaction with a merchant and you go onto a marketplace and you employ an AI to help you figure it out. They take out a percentage of whatever they recover, while you’re off at the beach or at your job. The results are delivered to you after the AI has negotiated the settlement. That is something that I think is going to be widely available.
Michael Moran: As an investor, what do you see right now that is attractive and either means quick wins or long holds?
Barbieri: Well I think in the end, the subject matter that we’ve been discussing today is that there are long holds. As bullish as I am about things like autonomous vehicles, I think we can see that national-level deployment of complex systems that involve interfaces between software and hard earth are, in fact, messy.
Human nature is one that tends toward entropy. And it’s hard to overcome that entropy with a system that may be perfectly created in a lab, but is difficult to deploy out on the streets.
As good as the technology is, it is still being improved and adopted. There is this very real hurdle of deployment in real world situations that I think a lot of my peers tend to overlook. So I would think the timelines are far more extended than I would care for. That is essentially inevitable.