At the recent TechSurge Deep Tech Summit, Vice President and Chief Internet Evangelist for Google, Vint Cerf, sat down with McKinsey Senior Partner, Lareina Yee for a wide-ranging fireside chat discussion on current tech trends.

You can read a brief transcript of part of the discussion or watch the full interview with Vint below.

‍

Q:  Reflecting back, what are some of the positive and negative unintended consequences of Internet technologies?

A: All powerful technologies have the unfortunate possibility of being used in a bad way. I wish there were ways to counter some of it. Back when the internet was created, you were pariah if you sent spam messages and, now, we all get dozens of unwanted messages every day.

On the other hand, though – wow, this has turned out to be amazing. The things that we can do by hooking computers together and hooking people up to each other are astonishing. In fact, many entrepreneurs and venture capitalists or investors would not be where they are today if it weren't for this virtual ability to hook people together and empower them with computing technology. It's how things work today. We've all learned lessons. Despite all the stuff that you could point out and say, boy, I wish that didn't happen, I would argue that mostly we benefited from all this.

Both the engineering side and the entrepreneurial side of AI owe the general public something: be thoughtful about what it is that you might be unleashing. We can't always predict what's going to happen with these technologies and to be honest with you, most of the problem is people.

‍

Q: What are your thoughts on the progress of AI? 

A: I want to be careful to distinguish artificial intelligence from machine learning, which is what we really are doing, machine learning. It has a variety of ways in which you can do it. I think we're seeing very powerful ways of using correlation, which is what a lot of these engines do, and why machine translation has worked out well, not perfect, but remarkably well. In addition to just literally doing either the transcription or the translations. So those things that depend on strong correlation to work well.

On the other hand, what everybody's talking about is Chat GPT or Bard. We know that it doesn't always work the way we would like it to. What it knows is how to string a sentence together that’s grammatically likely to be correct and it’s drawn from the material that it was learned from. So, the fact that it is based on factual material doesn’t help because of the way it produces text.

We should be thinking hard about how to cope with that because the system itself has no clue about what it's doing, as I see it. It's just generating likely sentences from the preamble that it's produced so far. The problem is you can't tell the difference between an eloquently expressed lie and a factual statement. We are a long way away from real conscious intellect.

What we're seeing is people trying to retrain these systems to be less likely to say something wrong, I think we're still a long way away from being able to do that successfully. Because the rate of human feedback is very low compared to the amount of material ingested into the systems for training.

‍

Q: What would you tell business executives who are considering implementing generative AI in their companies?

A: There’s an ethical issue here that I hope you will consider. On the engineering side, I think engineers like me should be responsible for trying to find a way to tame some of these technologies so that they are less likely to cause trouble. Of course, depending on the application, writing a not very good fiction story is one thing. Giving advice to somebody that has medical consequences is a very different thing. Figuring out how to minimize the worst-case potential is very important.

Both the engineering side and the entrepreneurial side owe the general public something: be thoughtful about what it is that you might be unleashing. We can't always predict what's going to happen with these technologies and to be honest with you, most of the problem is people. They will often seek to do that which is to their benefit. We have to remember that, and we need to be thoughtful about how we use these technologies.

Q: What advice do you have for entrepreneurs in business?

A: Staying focused is super important. There's such a thing as insurmountable opportunities where there's just so many things you could possibly do, you can even figure out how to do it, and yet you can't really afford to do them all.

I think the other thing I would say is that if you discover something isn't working, it's a good idea to discover that as soon as possible and pinpoint whether you can either make it work or do something different. I like projects that involve patience and persistence. Those two things really count. But of course, you must have the resources that let you exercise that patience and persistence.

‍

[Discussion transcript slightly edited for brevity and clarity.]

‍