The story of AI told by the people who invented AI


Welcome When i was thereNew oral history project from On machines we trust Podcast. It features stories of how artificial intelligence and computer breakthroughs came about, as told by those who witnessed them. In this first episode, we meet Joseph Attic, who helped create the first commercially viable facial recognition system.


This episode was produced by Jennifer Strong, Anthony Green and Emma Cillekens with assistance from Lindsay Muscato. Edited by Michael Reilly and Mat Honan. Mixed by Garrett Lang, sound design and music by Jacob Gorsky.

Full transcript:


Jennifer: I am Jennifer Strong. On machines we trust..

Here I would like to talk about something that we have been working on for a while.

It is called When i was there..

This is an oral history project featuring stories of how breakthroughs in artificial intelligence and computing came to be… as those who witnessed have said.

Joseph Attic: When I walked into the room I found my face, pulled it out of the background and said “I can see Joseph”. It was the moment when I had hair on my back… I felt like something had happened. We were witnesses.

Jennifer: We start things off with a man who helped create the first commercially viable facial recognition system… in the 90s…


It’s Joseph Grenier. Today, I am president of ID for Africa. It is a humanitarian organization focused on providing digital identities so that Africans can access and exercise their rights. But I am not always in the humanitarian field. After earning a doctorate in mathematics, he made some fundamental advances with his collaborators, which led to the first commercially viable facial recognition. This is why people call me the founder of the facial recognition and biometrics industry. When I was at the Institute for Advanced Study at Princeton, doing research and mathematical research, an algorithm was revealed on how the human brain recognizes familiar faces. .. But no thought was given to how to implement such a thing.

It was a time of programming and failure, and a long month of programming and failure. And one night early in the morning, indeed, we just finished a version of the algorithm. I sent the source code for compilation to get the executable code. And we got out and I went out to go to the bathroom. And when I got back to the room and the source code was compiled by the machine and sent back. And usually when I compile it it runs automatically, and when I walk into the room I find someone walking around the room, find my face, snatch it from the background, and say ” I see Joseph ”. And that’s when the hair down her back came in – it was like something had happened. We were witnesses. And I started calling other people who are still in the lab, and each one of them will bring them into the room.

And he would say: I will meet Paul, I will meet Joseph. And I took turns running around the room to see how many I could find in the room. Theoretically, no further breakthroughs were necessary, but it was a moment of truth, and years of work finally resulted in a breakthrough. The fact that I figured out how to implement it and finally confirmed that the feature worked was very, very gratifying and satisfying. We have developed a team which is a development team, not a research team. The team focused on integrating all of these features into the PC platform. And that was the birth of 1994, and indeed the birth of commercial facial recognition.

My concerns started immediately. With cameras more and more prevalent everywhere, the commoditization of the computer and computer processing power improving more and more, we saw a future where there was no place to hide. And in 1998, I told the industry that I had to put pressure on the industry and come up with principles for responsible use. And we felt good for a while because we felt we got it right. I felt that I had introduced a code of responsible use and then followed regardless of the implementation. But the code couldn’t stand the test of time. And the reason behind this is that we did not anticipate the emergence of social media. Basically, when we established the code in 1998, we said that the most important part of facial recognition systems is a tagged database of known people. I said the system would be blind if I wasn’t in the database.

And it was difficult to build a database. We had to scan each image and manually capture it, which allowed us to create up to 1,000, 15,000 and 20,000. In today’s world, the beast can be out of the bag. Help him by feeding billions of faces and tagging yourself. Well, we are now in a world where it is difficult to hope to control and demand that everyone be held accountable for the use of facial recognition. At the same time, you can just rub, as is the case in some companies these days, so that you do not run out of familiar faces on the Internet. So I started panicking in 2011 and wrote an editorial saying it’s time to hit the panic button as the world moves towards ubiquitous facial recognition and face availability everywhere. In the database.

Back then people said I was vigilant, but today they are aware that this is exactly what is happening today. So where do we go from here? I lobbied for legislation. I have been lobbying for a legal framework in which you are responsible for using someone’s face without their consent. And this is no longer a technical problem. This powerful technology cannot be contained by technical means. We need some kind of legal framework. Technology cannot get too far ahead of us. Beyond our values, ahead of what we consider acceptable.

Consent issues remain one of the most difficult and difficult issues in technology, and telling someone is not enough. I have to give my consent. They need to understand the consequences of what this means. And not just say, well, we signed up, and that was enough. We talked to people, and if they didn’t want to, they could go anywhere.

It is also easy to be fascinated by the showy technical features which can bring short term benefits in our lives. And we realize that we have given up something too precious. And at that point, we were desensitizing the population and we were unable to push them back. This is what worries me. I’m afraid facial recognition may be possible with works like Facebook and Apple. I am not saying that everything is illegal. Many of them are legal.

We’ve gotten to the point that the general public sees it everywhere, which can be secular and insensitive. And maybe 20 years from now you will be leaving your house. You no longer have the hope that you won’t. He is not recognized by dozens of people crossing along the way. At this point, I think the public will be very careful as the media start to report cases where people have been harassed. People were targeted and people were selected and kidnapped based on their net worth on the streets. I think that’s a lot of responsibility in our hands.

So I think the issue of consent will continue to plague the industry. And that probably won’t be resolved until the issue comes to fruition. I think we need to set limits on what we can do with this technology.

My career has also taught me that, as we know today, facial recognition was invented in 1994, so it is not good to go too far. But most people think it was invented by Facebook and machine learning algorithms. It is currently growing all over the world. I had to step down as a public CEO at one point because I was cutting back on the use of company-promoted technology for fear of harming humanity. Scientists therefore believe that they must have the courage to plan for the future and see the results of their work. I’m not saying they should stop making inroads. No you have to do your best and make more breakthroughs, but we’re honest with ourselves and basically let’s do the world and politics that there are positives and negatives to this breakthrough. Need to warn people. Therefore, using this technology requires guidance and a framework to steer it towards positive rather than negative applications.

Jennifer: When was I there … Is an oral history project featuring stories of people who have witnessed or created breakthroughs in artificial intelligence and computing.

Is there a story to tell? Does anyone know Please email [email protected]



Jennifer: This episode was recorded in New York in December 2020 and was produced by me with help from Anthony Green and Emma Cillekens. Edited by Michael Reilly and Mat Honan. Our sound engineer is Garret Lang… Sound design and music by Jacob Gorski.

Thanks for listening, Jennifer Strong.


Leave A Reply

Your email address will not be published.