Here is an audio interview with David Riecks about Artificial Intelligence (AI).
What is AI? What is the digital and/or fourth revolution about? What does AI mean for me? My future? My children’s future? The future of humanity? Should I be scared? Worried? Excited? Will I pee my pants a little? Welcome to AI Fears. Join us as we explore the answers to these questions with interviews, panels, and discussions. Developments in AI are going to impact humanity. We think we all should be a part of it, or at the very least understand what is going on.
Henrik de Gyor: This is AI Fears. I’m Henrik de Gyor. Today I’m speaking with David Riecks. David, how are you?
David Riecks: I’m doing okay.
Henrik de Gyor: David, who are you and what do you do?
David Riecks: Well, my name is David Riecks, and I wear several hats. I started out as a photographer and got heavily involved in digital asset management in the early 1990s. Based on some of the work I had done there, I was asked to join the IPTC working group that was integrating the Adobe XMP into the older IPTC standard. And along the way helped to launch updig.org, photometadata.org, and the embeddedmetadata.org websites. And in addition to that … kind of along the same times, I got involved with keywording images because I needed to find them later, and this eventually became controlledvocabulary.com and resulted in a set of hierarchical keywords that you can import into a dozen or so applications like Adobe Bridge, Lightroom, Photo Mechanic, etc.
Henrik de Gyor: David, do you think a robot will take your job?
David Riecks: In the near-term, probably not. Although what I would say is whatever job it might take of mine, and I’ve got several, I think I might see it coming and either figure out a way to exploit it or be moving onto something new by that time.
Henrik de Gyor: Do you think we need ethics in AI?
David Riecks: Oh, without a doubt. There’s actually an academic discipline I was reading about not that long ago just within the ethics area which addresses this growing need. It’s more than … I don’t know if you’ve ever heard of the trolley scenario, but it’s a lot more than that for driverless cars. So imagine all the problems we’re having with drones and there’s legal issues but there’s also … I don’t think anybody wants to envision a future where we’ve got all these buzzing around all over the place. There needs to be some rules and drones themselves if they’re autonomous need to respect us. So, part of the ethics gets into the inherent biases in all these systems because they’re made by humans.
Pretty much every human whether they want to admit it or not has some baggage in their life, so when they’re writing or programming or whatever those biases are going to sneak in, there’s just no way to avoid it.
There was something on NPR not that long ago where they were talking about this Canadian Psychologist who has developed a test, like a psychopathy checklist I think is what they called it. But it was basically to figure out, are people potential psychopaths? And other things like that, and somehow it got way late into being used to determine whether prisoners could get parole or not. So, those are ways that things could get messed up.
Henrik de Gyor: Based on books, media, movies regarding robots and artificial intelligence, what’s your favorite doomsday scenario?
David Riecks: Favorite Doomsday, that’s kind of an interesting way of phrasing it. I mean, Skynet becoming self-aware in the Terminator series is pretty frightening. But did you ever see the British TV movie called The Last Enemy?
It had Benedict Cumberbatch in it, that scenario is a lot more likely in the near-future. Basically having your identity revoked. You wouldn’t be able to buy food, take a cab or public transport. That’s a lot more realistic and pretty frightening.
Henrik de Gyor: And what’s your favorite bright-future scenario?
David Riecks: Bright-future, you know I had to think about that for a while. There was a book I read awhile back by Robert J. Sawyer called … it was a part of trilogy, they all started with www, and the first one’s wake and it basically talks about how an AI comes to life. And the other books kind of traces what it ends up doing and making the world a better place. Probably more people have seen Her, the Spike Jonze movie with Joaquin Phoenix, and in that one they do kind of make the world a little better but then they decide to leave. And we’ll just work it out on our own. By in large, most science fiction is pretty heavily into the dystopian side of things.
Henrik de Gyor: David, do you feel secure about your job today or in two years, five years, ten years, in 20 years?
David Riecks: Today, I’d say that’s pretty secure. Two years… kind of reminds me of the sign you always see in the sideview mirror “objects that were closer than you think.” Five years you know I think I’d pull out my magic eight ball and find a hazy “try again”. But ten years, 20 years down the road I think we’re at signs point to “yes” and I’d be surprised if it was only a vestige.
There are so many things that don’t get completely replaced, but they get sub-snooted in the whole thing.
[Looking at] Film photography in the transition to digital, but you’ve even got things like cobalt programming, but that was I think first 1950’s or 1960’s and they expected it to be around for ten years and I know there are a lot of insurance agencies and places like that, that are still using them.
Henrik de Gyor: David, we’re living with a prediction that in the next decade, fifty-percent of jobs will be removed from the job market.
How do you feel about this, do you believe it’s true? Why or why not?
David Riecks: Well, I think where when you say removed … I mean jobs will certainly change, and there a lot of disruptive events that have occurred in the last 20 years. And at the rate at which they are being disruptive, this is probably only likely to increase.
Maybe exponentially, maybe not? I have a nephew, he just started college last year and he and I have talked about this issue for the last couple years. He wants to get a good job and he doesn’t want to see it disappear before he’s done paying off his student loans. But, much of what happens is an evolution rather than a revolution, so if you look at some of the AI, you look at voice recognition, OCR’s and it’s been around nearly as long as- But it’s still, I think at best something like 90 to 90 [percent accuracy].
Which sounds good because that still means like eight or ten words out of a hundred are still wrong. And we look at cars, people thought we’d have flying cars by now and we don’t really have that. But we do have cruise control and automatic parallel parking, and driverless cars are just around the corner, and it’s probably going to be that way for a while. It’s going to be phased to we’ll be still sitting in the car using autopilot, and they’ll be able to do some things but the AI is not going to be present there.
Henrik de Gyor: Based on your experience, what is the biggest challenge and success of AI?
David Riecks: Just talking about the optical character recognition, voice recognition has gotten pretty good as well but fisher in her Midwest accent like I am. I know I’ve heard horror stories about Siri with Scottish accents, it still has some problems with that.
Language translation’s something that’s kind of interesting as well. I think that’s … has to be 100 percent accurate for you to get a good idea of what people are talking about in another language and it allows people to correspond with each other.
I think we still have a ways to go with image recognition and video recognition is kind of hot on its heels as well. If you look at maybe not so much a success … there was a thing a couple of years ago called Galaxy Zoo, it’s a crowdsourced astronomy project and they invite people to assist in classifying galaxies. Computer programs at least up until that time had not been able to reliably classify this images, and/or took a lot of time and when they crowdsourced it, people were able to zip through them. So they’re maybe somethings that humans are still better at than computers.
Computers will work for hours and hours with no coffee break.
Henrik de Gyor: It’s true, I think you maybe still training the AI, that’s your point. The humans still have to train the AI with or without biases to your point.
David Riecks: Yeah. There’s a good job for the future.
Henrik de Gyor: True. That will take a while.
David, when thinking of robots in AI, what are you excited or downright afraid of?
David Riecks: I think I would more look at the excitement part, there are some certain problems that need to be dealt with. But what we’re looking at right now within the image area is being able to leverage AI to manage large amounts of rich content.
So for instance, you’ve got compute power becoming available on the cloud to help automate metadata generation. So we’ve got Google, Amazon, Microsoft and others releasing these API’s that will allow you to send images through keywords and confidence level back. That’s all well and good but what I’d really love to see is leveraging maybe some of that voice recognition to allow me to talk into my camera and have that image embedded into the images so that they fly off with them. When they go into any kind of asset management system, they automatically have a caption and where it was taken.
You can get some of that from GPS, but you can’t necessarily have GPS tell you what was in the frame at the time you were taking the picture.
Henrik de Gyor: That’s true. I’m shocked that camera manufactures haven’t figured out how to add that onto the audio part that you were just mentioning or the image recognition part. Because right now, cameras to point only recognize distance and light, right? Aside from GPS, but if they recognized what was in the subject it could add a lot more value to it is your point?
David Riecks: Well, some of the newer cameras can do facial recognition where they can tell if there’s a person seen, they’ll draw a box around it. But it doesn’t know who that person is.
Henrik de Gyor: Yeah, no that’s another step.
David Riecks: But even just being able to tag stuff with it is not [there].
Henrik de Gyor: David, what advice would you like to give to those fearful of losing their jobs?
David Riecks: Well what I’d recommend is read and learn as much as you can. Take things apart, put them back together. Understand how things work.
When I started in Photography, everything revolved around pinnacles … and in the span of 10-15 years maybe, it transitioned to computers, digital pixels, sensors … there were lots of interim technology. scanning, prints, photo CD, now we’ve got camera scanning. But in all of that, there was generally some knowledge that you could leverage. But if I just sat back and said “ah, it just a phase. Chemical photography is still going to reign.”
I wouldn’t be where I am now, I’d likely be jobless, bitter and reading lots of dystopian science fiction.
Henrik de Gyor: David, what are your hopes for Artificial Intelligence?
David Riecks: Well … I think we talked … the other one I had left, we were talking about what I’m excited or downright afraid of I thought.
Henrik de Gyor: Sure, please.
David Riecks: But maybe we could switch those around?
Henrik de Gyor: Yeah, I apologize I did.
David Riecks: Overall I’d have to say I’m excited about what AI can bring, but I do think as Asimov goes on to something that is three laws of robotics, I don’t know if you’re familiar with those?
In fact, those were kind of the inspiration for the original principles when I was writing the Metadata Manifesto was trying to get it down to three core nuggets. And that was a decade ago, that eventually became the embeddedmetadata.org website and I think we expanded it out to five. But trying to keep something simple, having checks and balances built into the system is a good idea.
Henrik de Gyor: Great, well thanks David!
David Riecks: No problem, thanks for having me.
Henrik de Gyor: For more on this, visit aifears.com
Join us regularly for topics and posts on the current state of AI.
Subscribe and participate in the discussion.