• 5 months ago
Vinod Khosia, Chairman of Khosia, spoke at the Imagination in Action's 'Forging the Future of Business with AI' Summit on why AI will be in everyone's life in the near future and why robotics will be the next sector to have its own AI moment.

Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1

Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:

https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript

Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com

Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.

Category

🤖
Tech
Transcript
00:00 Let me ask you, what was it like being the first investor in open AI?
00:06 You know, were you just putting money on 21, let it ride?
00:10 Did you have insight that you thought it would be big?
00:12 Is it what you expected?
00:14 And what can you tell us as an audience, as you've seen that kind of evolve, that we should
00:21 know that only you know?
00:24 Well, the thing to realize is you have to have conviction about what you're doing.
00:30 These things, especially looking into the future, as Yogi Berra said, is very hard to
00:36 predict.
00:38 But you have to have conviction that the directional trend going on, that's important.
00:44 I've long believed in AI as a directional trend, wrote about it in 2012, taught it'd
00:52 be important, and hence when the open AI opportunity came about, which was we made the decision
00:57 in 2018 to invest twice the amount I'd ever invested in any other company as a first investment.
01:09 And that was because I had conviction about what AI could do.
01:12 I had no idea what the timing would be or when the breakthrough moments would come.
01:16 But directionally, I felt there was so much talent going into AI, it'd be really, really
01:21 exciting.
01:22 Vinod, will we be capable of having free AI doctors for every person, AI tutor for every
01:31 child 7 to 11, you know, that store 7 to 11?
01:36 No, no, 24/7.
01:39 No question about it.
01:41 You know, this is not even a question anymore with the capabilities of AI.
01:46 We have to go through the full process of FDA approval of an app as a primary care physician,
01:52 which sounds crazy to people, but almost certainly can happen if the FDA lets it happen.
02:01 No question if we do a one-on-one comparison of 10,000 encounters between humans handling
02:09 it and then AI handling it, the AI would do substantially better, especially the more
02:14 diverse the community, the most diverse the input cases.
02:18 That will absolutely be true.
02:20 Vinod, will we have a billion bipedal and other robots freeing humans from the servitude
02:28 of undesirable jobs in the next 20 years, or is that not going to happen?
02:34 I believe robotics is coming up with a chatty-pitty moment here pretty soon in robotics in the
02:40 next four to five years.
02:43 Could be as soon as two or three years.
02:48 Robots aren't programmed anymore.
02:50 They are learning systems.
02:53 And I think that'll be the seminal change in robotics.
02:56 And they understand the physical world.
02:58 They understand physics, real-world dynamics, all that kind of stuff.
03:05 And because of that, I'm actually pretty optimistic that many of the things that have looked very
03:11 stunted and terrible because we as humans are trying to program them, and those are
03:16 like rule-based AI systems of old, we'll see this rapid change.
03:25 So I'm pretty optimistic we'll see that happen.
03:28 And then nobody has to work at a GM assembly line for eight hours a day for 40 years.
03:37 That's not a job.
03:38 That's servitude.
03:39 So Vinod, did you ever use punch cards to program?
03:44 I did start on punch cards.
03:45 I started my programming career in India on punch cards, so in 32-kilobit machines.
03:53 You could probably guess where I'm going with this.
03:55 Will there be a billion-plus programmers all programming, not in punch cards, but in human
04:01 language dramatically increasing the scope of computers?
04:04 Will computers adapt to humans, not humans adapt to computers, like the Palm Pilot?
04:10 They couldn't get the Newton to work to read your handwriting.
04:12 So Jeff said, "Hey, let's just have people adapt."
04:16 Are we going to have a tipping point where we flip back?
04:22 I absolutely think so.
04:24 Those are two entirely different concepts, but let me explain them just a little bit.
04:32 First, certainly, people will be able to program in natural language.
04:41 And so many more applications.
04:43 Essentially, programming will also be near free, and many more people will be able to
04:49 do it.
04:50 Hence, the use of computers will expand dramatically.
04:53 You might actually do use one software because it's so easy to do.
04:59 It's your dispatchable or throwable packaging for your shampoo bottle.
05:05 You throw it away after use.
05:09 I think computing will become more like a utility, more like electricity in the background.
05:16 Most people, most of the time, won't be sitting in front of computers.
05:20 Hopefully things will happen for them.
05:23 I'm pretty optimistic about that.
05:25 The other major change we will see is, so far, we've had to learn how to insert a row
05:35 in Excel or use the menu to do something or, God forbid, the macro programming language
05:41 in Excel or learning to use SAP.
05:45 Computers will learn humans instead of humans having to learn computers.
05:49 I think that'll be another major trend.
05:52 Devices like the Rabbit, you can talk to and they do things that you need done.
05:58 They understand it, can break it down into the various tasks that might be needed.
06:04 I'm pretty excited about that.
06:07 Computers will become a lot, lot easier to use.
06:11 Vinod, do you get out to movies?
06:15 Do you like entertainment?
06:18 I love entertainment and design.
06:20 I know where you're going.
06:22 Yeah.
06:23 So, my next question is, how will AI play a role in entertainment and design?
06:31 Some people say, the Beatles won't happen and then this next wave of AI because you
06:38 don't need a group.
06:39 You can use AI to do it.
06:40 Will the human collaborate with the AI and be creative?
06:44 What's the future of entertainment design?
06:46 Is that a dead-end opportunity or is that just getting started?
06:51 It's just getting started.
06:53 Five years ago when I first talked, I gave a talk in Toronto, I gave a talk about AI
07:01 generating music.
07:02 It was met with real skepticism, even derision.
07:07 Not only will it happen, but also that we don't need it.
07:16 I think this notion some techno haters have of some human elements will sort of disappear,
07:29 I think.
07:30 I think AI alone or AI+ in conjunction with humans will have much more creativity.
07:39 We'll see much more what sometimes is called soul in the music.
07:46 We will also see much more diversity of entertainment, not only music, but everything else.
07:53 I think broadly, we will see much more content, near free, much more available, much more
08:01 personalized.
08:02 You might get music for the mood you want to be in just before you give a talk or just
08:07 before you are at an athletic event.
08:09 You might get very personalized music for your brain to you.
08:17 I don't want to say the celebrity fan relationship is going to change.
08:22 Taylor Swift will still be there and Swifties will still be there, but that doesn't take
08:28 away.
08:29 So this AI music doesn't take away from that relationship and that kind of performance
08:36 or entertainment.
08:37 Next question is just a one word answer.
08:41 Are you a Swifty?
08:42 No.
08:43 Okay.
08:44 All right.
08:45 Internet, agents, access.
08:48 What comes to mind on those words?
08:51 You know, I think most use of the internet will be by agents.
08:59 It won't be by humans.
09:01 24/7, there'll be billions and maybe tens of billions of agents running around, multiple
09:09 ones for each of us, doing specialized things for us.
09:14 Almost if each of us had multiple assistants to do things for us.
09:21 So most internet access, in my view, will be by agents, not by human beings.
09:27 That might have some implications on how content on the internet needs to change.
09:35 So Vinod, I'm curious, open versus closed.
09:41 What are your thoughts on that?
09:42 And are you passionate about one way or the other?
09:47 Has your opinion changed over the years on that?
09:51 You know, so my opinion hasn't changed a lot.
09:55 You know, Sun was the first really open source company.
09:58 NFS was the first major product you think of, even before Linux.
10:04 So that's pretty exciting.
10:10 You know, we invested in GitLab.
10:12 So I've been a big fan of open source.
10:15 Now in the case of state of the art models, I feel like open source shouldn't be prohibited.
10:22 It's like open sourcing the Manhattan Project, because we have an adversary in China.
10:27 And I believe the AI race is so important.
10:31 Whichever country, whether the West wins it or liberal value countries build it more generally,
10:39 because I consider many, many countries around the world, liberal values companies, or China
10:45 wins it will have huge implications for the planet.
10:48 And I hope we win it.
10:51 And hence, we shouldn't be open sourcing state of the art models.
10:55 Everything else is fine.
10:57 In terms of AI, what are you most excited about?
11:01 And what keeps you up at night and you're most worried about?
11:06 You know, the hard thing in AI is to predict what capability will show up.
11:14 GPT-5 hopefully is coming soon.
11:17 What will GPT-6 be like?
11:19 When will GPT-X start training X plus 1?
11:25 Those are a set of things to worry about.
11:27 It impacts the rate of change.
11:31 I wonder if once we have AI programmers like Devin, are we going to see really, really
11:36 interesting progress.
11:38 But at some point, these AI systems will be designing other AI systems.
11:42 And then you have real issues there.
11:48 That's on the model side.
11:51 I'm also pretty optimistic about something most people don't believe, that within the
11:58 next two years or three years, we'll start to see more interesting alternative approaches
12:08 that add substantially to the LLM model is the right way to say.
12:15 So you might see other approaches like probabilistic programming.
12:20 There's a group at MIT very interested in that.
12:23 You might see symbolic logic.
12:25 I'm pretty excited about a company we have called Symbolica that's doing something called
12:30 category theory.
12:32 In fact, it's funny, nobody's heard of it, but I've run into two or three startups doing
12:36 category theory, including one at MIT.
12:39 So these are esoteric areas.
12:43 That's sort of what OpenAI might have looked like in 2018.
12:50 So the theory is there, the attention paper was out, and then you didn't see its implications.
12:57 So I think they will all be additive to current models.
13:02 So that's sort of the model side, I think, will continue to evolve pretty substantially.
13:09 Then we have the application side, and we've sort of talked much about it.
13:15 There'll be a lot of copilot type applications.
13:20 Almost everybody should be able to use a copilot for most things they do.
13:28 Why a physician would ever have to use Epic, I don't know.
13:31 They shouldn't need to.
13:33 And AI will do that for them.
13:37 But areas, and there'll be lots of new consumer applications, how shopping is done.
13:43 We've talked about music or games will all be impacted.
13:47 There'll be regular enterprise.
13:49 But the area I'm most excited about is when you replace human expertise.
13:55 We've talked about near-free doctors.
13:57 We've talked about near-free tutors.
14:00 I think we'll have near-free oncologists, structural engineers, you name it.
14:08 So expertise, we'll have geologists, so we can discover a lot more mineral resources.
14:13 That's an area I'm very interested in.
14:16 So that side will also go through a lot of innovation.
14:21 In fact, I think a lot of media will be reinvented too.
14:25 We have a book publisher that uses AI, and they can, they're 10x better than the best
14:31 book publishers in picking out hits that might be million-dollar wins, which is a good level
14:39 for a book.
14:40 So Vinod, what advice do you have for college students who are thinking of dropping out
14:46 or maybe finishing on areas to focus on?
14:49 Is software engineering dead?
14:51 And what skills are needed?
14:54 I don't think it's fair to make any assumptions.
14:59 I think the best software engineers will become 10x or maybe 100x engineers.
15:03 Or they'll still be needed, but they'll be a lot more capable or powerful, amplified
15:10 by AI.
15:13 Simple programmers will become better programmers.
15:17 What I would say in general to young people is be as diverse as possible in the things
15:22 you learn.
15:23 Don't plan on learning your career now, but do the courses that help you learn how to
15:32 learn so you can adapt rapidly.
15:34 The one thing for sure, we will have to keep changing what we do.
15:39 Nothing we learn today in class will be relevant 10 years from now.
15:44 So mostly what you have to do is learn to learn and be able to change fields rapidly.
15:49 Different advice for PhD students.
15:53 That's a very different thing.
15:57 Let me ask you this.
15:58 Did you know that there was a bored kind of Shakespeare drama at OpenAI?
16:01 Were you aware of that?
16:03 No, I didn't hear about it.
16:06 So this is an intimate conversation you and me here.
16:10 Do you want to say anything, weigh in on that, that no one knows what really happened or
16:14 went on?
16:16 Well, the only comment I would make in general is the broader issue is one of how you handle
16:24 risk.
16:25 AI systems do have risk, but we have lots of risk as society.
16:33 An asteroid could hit the planet too.
16:35 That's a real risk.
16:37 Another really big risk is another pandemic.
16:40 So humanity faces a whole basket of risks.
16:45 And so you learn to manage them, you learn to pay attention, you don't go overboard on
16:49 one thing, especially, you know, EA was a Jonestown like religion.
16:55 It's just completely irrational, it pretended to be rational, it's completely irrational.
17:03 And so ignoring every other risk in humanity, ignoring all the benefits of humanity, it's
17:12 just silly.
17:16 How do you like, how do you use AI?
17:18 Like when you wake up in the morning, you know, does it brush your teeth?
17:22 You know, are you an AI right now?
17:24 Do you use AI to, you know, how are you using AI?
17:29 And how has that changed from a week ago, a month ago, five years ago?
17:35 So before I go, I would say to you, I think we have to take a long view of AI, even AI
17:43 applications and the silly, you know, you have to realize the press needs headlines,
17:49 so they have to make headlines.
17:51 And they have to show how silly X or Y is, or why this hand that Midjourney produced
17:58 has the six fingers, not five.
18:01 By the way, nature does that sometimes too.
18:05 But we have to take the long view.
18:07 All of these systems are so new, version 10 of them will be mature, or somewhere near
18:14 what we might consider mature.
18:17 And so we have to be patient and say they're good at some things, they're not good at others.
18:23 And what we have to do is use them enough, start the flywheel, data flywheel, and what
18:29 doesn't work, so we can fix it.
18:32 And every generation, every year will be better than the last.
18:35 And that's what I'm optimistic about.
18:37 The rate of growth will become exponential.
18:40 Systems that evolved to be better are almost always better than engineered systems that
18:45 are precise in one dimension.
18:48 And once AI gets to be really learning systems, which is, it isn't quite today, I think we'll
18:54 see even faster progress.
19:00 So my last question to you, Vinod, is are you having fun?
19:07 I'm having a blast.
19:08 I mean, how can it not be fun working on all the things we've just talked about?
19:12 And I work on those, and a lot of other fun things like fusion and Mach 5 flight and all
19:18 that kind of stuff.
19:20 Well, thank you very much.
19:22 Yep, ladies and gentlemen, you've been Vinod Khosla.
19:24 Vinod Khosla.
19:26 Thank you.
19:27 [APPLAUSE]
19:27 [END PLAYBACK]

Recommended