Off the Editor’s Spike: ‘Deep fake’ and neural what?

It was quite a gathering last Saturday for mom’s 92nd birthday. A whole new crop of great-grandchildren littered the floor, to the delight of everyone. It was a good day. A very good day.

I noticed mom’s cup of coffee was cold. As I maneuvered through the babies, toddlers, diaper bags and blankets to reach the kitchen for a refill, I entered a different world.

I was looking for the coffee pot but in a very real sense, the few crowded between the stove and refrigerator might as well have been standing at the water cooler during a break at some large corporation. The discussion was intensely focused on Tesla and a future defined by artificial intelligence. I couldn’t help eavesdropping, thankfully arriving at the conversation much too late to do anything but listen.

The thirty-somethings were exchanging views on where unharnessed artificial intelligence would lead our society. There was debate on competing ideas. There also was agreement that realistically, nothing was going to stop AI from encroaching on our present way of life.

The conversation probably started by some comment about the autopilot computer vision systems now released for self-driving automobiles. This led to a discussion about neural net applications. Understand, the emerging AI technology is already making

website pages and smartphone apps obsolete.

Neural net programming is called deep learning. This enables machines to emulate how a human brain works. Thus, computers can learn. The discussion I overheard was about a breakout announcement by Tesla anticipated later this year that promises to vault AI to incredibly new levels.

One young father prefaced a remark about the future by alluding to holy scripture. Then he relayed how some experts see a future where AI will be in conflict with mankind, a real life doomsday script similar to many sci-fi movies.

I didn’t hear these young adults embracing such predictions. But they were definitely giving it serious thought. That startled me. But then, I’m a greybeard who has trouble using a smartphone and feels somewhat challenged to deliver a hot cup of coffee to his mother without spilling it.

The discussion got me refocused on another high tech topic which prompts nearly identical anxieties: deep fake videos. What I read reveals it to be the next threat to truth. And we aren’t ready.

We may soon face totally convincing videos showing events that never happened — created so effectively that even experts will have trouble proving they’re fakes. Deep fake video will show people saying things they never did with the authenticity of their own voices. We’ll watch images showing actions that a person never did, by blending their images with other video or creating new images entirely from scratch.

This is dangerous. It’s much more serious than showing a football team scoring touchdowns that never happened. We live in a society that uses video for credibility — to learn what really did happen inside our school buses, to what really happened in the parking lot, to what happened outside the patrol car on a traffic stop.

Think on this. Now society will have to learn that video no longer guarantees reliability. Instead, video could be the biggest lie of all.

Most fake videos in use right now are still far from perfect. But things are changing. Just compare the video quality of the initial version of Madden NFL Football released in 1988 to today’s video game with real player motion. The kids understand … even expect and accept such change without considering all the consequences.

I stood there and listened to the tech talk being shared in mom’s kitchen long enough for her fresh cup of coffee to get cold again. I had to concentrate to follow the fast-paced bantering about AI and neural networking. Getting old bothers me. But since then, I actually feel encouraged.

It is healthy — no it’s necessary — for more conversations on topics such as these to be voiced around the supper table in kitchen conversations throughout America.

There are experts looking at search engines and social media platforms to identify and expunge misconceptions and harmful applications of AI and fake video. Some alarms are being sounded in the technical, political and security worlds. But the dangers haven’t really gained sufficient public attention.

We have a window of opportunity, when verification procedures and ethics codes could still be put into play to mold what’s to come. We shouldn’t leave such important matters to others. We must discipline ourselves to individually learn and think hard and long about things we never heard of just a few years ago, like deep fake video and neural net.

Otherwise, I fear we’ll look back on these times of fake news as decried by President Trump and think, “How quaint.”