The Future of AI - What Will It Mean for Different Industries and the Job Market? with Amr Awadallah

Raja Iqbal:
Hello everyone, and welcome to Future of Data and AI. I’m your host Raja Iqbal. It is my pleasure to welcome Amr Awadallah to the podcast. Amr is a serial entrepreneur and a prominent name in cloud computing industry. We’ll talk to him about the journey as a technologist and his work at Yahoo and Cloudera, and his most recent startup, Vectara.

Raja Iqbal:
Welcome to the show.

Amr Awadallah:
It’s good to be here.

Raja Iqbal:
Okay, so which applications or industries you think will find it easier to adopt GenAI? and which ones it would be harder. Right. So, you know, there could be different use cases. There could be regulatory requirements, the quality of data and all that. That’s a a do you have any classification? or maybe based on your experience?

Raja Iqbal:
Which industries or which enterprises you thought. Yes, it is going to be a bit of a uphill battle for them to adopt, GenAI versus these are slam dunk easy use cases.

Amr Awadallah:
Yeah, that’s an excellent, excellent question. So first, in terms of which businesses and which verticals will be using this technology and will benefit from this technology, I usually answer the question by asking a question and I’ll ask you this question. Raja. Which businesses should have software as part of their business?

Raja Iqbal:
Every.

Amr Awadallah:
That’s exactly I mean, every business needs to have software business. And why is that? Because software is adding a lot of efficiency. It’s doing it’s automating a lot of things, making them happen a lot faster. And unlike humans, when you figured out how to do something with software and it can make mistakes. But once you fix that mistake, that mistake never happens again.

Amr Awadallah:
You write humans, you punish them. They make the mistake. a couple of years go by and they can make the same mistake again. Right? And that’s a very, very big problem with humans. And that’s why we have software everywhere, is because software is helping and streamline and increase the efficiencies of our businesses, whether we’re agricultural or manufacturing, pharmaceuticals, life sciences, even a tailor and accountants.

Amr Awadallah:
We all need to have software around us. same thing would be true for AI. Like we’re seeing the beginning of it right now. But I expect every single business, every single business and every single industry will be affected by this within the next five years. Like it will happen. And if they don’t evolve to, embrace this and adopt this and leverage this, if they stay afraid of it and stay away from it, I’m afraid they’re gonna fall behind.

Amr Awadallah:
The other businesses would win, and we have many evidence of that. If you look at the IT revolution and the businesses and the banks and insurance companies and even the e-commerce shops, whatever, that truly leverage technologies, they became the leaders of the world. And the ones they did not, they disappeared. The it’s not it’s not that they became smaller.

Amr Awadallah:
They literally disappeared. You see blockbuster, look at blockbuster versus Netflix, for example. These many, many like that. So, that’s my preamble is every single business and every single industry will need to pay attention to this. Now, the second part of your question was for which ones will it be easier or harder? I think that’s like, I know it sounds like a fishy answer, but my answer to that is up to you.

Amr Awadallah:
It’s up to the business. It’s literally up to the business. Some businesses will have very strong leaders that will make it happen right here, right now, even in a very, very regulated industry. Like if you look at JPMorgan, for example, which is one of the leading companies and leading banks in the US, but they actually have a very, very strong leader.

Amr Awadallah:
They just came in and said, well, are you going to adopt GenAI? left and right. We go, it’s going to be here. And they have it now all over the organization. Right? So even though they are in a very, very nebulous, regulated industry, they, they they are of course accuracy is very important for them, etc., etc.. but if I put that aside, like so that’s my message of what I’m telling leaders in these businesses, you have control of how fast is going to move.

Amr Awadallah:
It’s going to be hard. It’s like it definitely was hard for JP Morgan to go through all the regulations and approvals. But if you have the will and if you have that conviction that AI is going to be the accelerator for your business in the future, you can make it happen, right? So that’s my preamble. Now, of course, it’s a no brainer that the harder businesses, to do this for are the business where, one of two things that are happening, either it’s hard to get your hands on the data.

Amr Awadallah:
The data is fragmented, it’s all over the place, or you haven’t digitized yet. So your data is still on paper. It’s not in digital form. So that’s number one. And in fact, in all of our POCs with customers and all of our evaluations, the hardest part is always the data. And how can we get the data to flow in a smooth way?

Amr Awadallah:
That’s it. Once it flows in a small, smooth way, the AI part is easy. It works just like that. But getting the data in a clear form and getting the right data. Because as we all often say, garbage in, garbage out is the hardest part. So that’s number one. Number two is, yes, of course, heavily regulated businesses will have a lot more steps and groups they will need to jump through for approval process.

Amr Awadallah:
And that can delay things a bit. But it’s not something that should be, an impediment. It’s something you have to do anyway. Even if you do this three years from now, you still have to do it. If you do that five years now, you still have to do it. Compliance is very, very important for a reason is because you’re dealing with very, in the case of finance, you’re dealing with people’s money.

Amr Awadallah:
In the case of health, you’re dealing with people’s lives. So you have to have these compliance and governance rules which slow you down a bit. But that should not be the reason why you don’t do it. This should not be there is oh, I’m not going to adopt AI because I have too much compliance. Then again you’re going to fail.

Amr Awadallah:
You’re going to disappear in five years from now. I promise.

Raja Iqbal:
Yeah. That’s, that’s great. So, hypothetically speaking, maybe more of a rhetorical question, but you you’re right. So if I was able to get all the books on, as a general practitioner, my primary care physician. So I took all the medical books and maybe some records and all of that and put them in direct data.So, will doctors be out of job?

Raja Iqbal:
I mean, what would it look like?

Amr Awadallah:
as I mentioned at the beginning of our conversation, every single job will be impacted by AI significantly over the next few years. In ways that we cannot predict actually, as well. So, I mean, just look at the like before we go to doctors, I mean to look at some of the jobs that already have been replaced or are being replaced in front of our eyes by autonomous driving like cars that drive themselves, like in San Francisco.

Amr Awadallah:
Here we have Waymo, and you can grab the Waymo app, which is just like Uber. And, and you say, I want to go from year to year and a car shows up with no driver in it, and you just go in the car and you get out on the other side, and it’s the smoothest drive. You’re having a half way better than a real driver, way safer.

Amr Awadallah:
The first time you do it, you got to be a little bit afraid because it looks like there’s a ghost sitting moving the wheel, which is very freaky. But once you do the first couple of right. Wow, it’s actually a better experience. Like I’m actually safer inside cars like that. Now, will Waymo make an accident every now and then?

Amr Awadallah:
Meaning what? The AI equals somebody to die every now and then. Probably. But and same thing if we automate medicine with AI will that was that AI doctor every now and then make a mistake probably. Would they make the same volume of mistakes that humans are making? Not even close. Like today. We have 00,000 people that die in the United States because of bad drivers, right?

Amr Awadallah:
00,000 people die and using that technology, the cars that we have invented decades ago are still dying today because of humans not learning from the mistakes, because of humans answering their phone while they’re driving or drinking and driving or putting on makeup while they’re driving. And then they end up having this bad accident and I will make that disappear.

Amr Awadallah:
Same thing for doctors. And doctors can go in and they’re doing a diagnosis, but they just had a personal issue with their spouse. The spouse was shouting to them and blaming them, or one of the kids did something that’s made them fume just minutes ago, and now they’re doing the diagnosis. Do you think that’s going to be a good diagnosis now?

Amr Awadallah:
It’s not going to be a good diagnosis. Only the best, best doctors actually know how to compartmentalize and separate their emotions from their logic. But most humans know that the emotions would affect their analysis, would affect their perception. So to have an AI assistant side by side with their doctor that catches them when they do that, say, no, actually this is not the right diagnosis.

Amr Awadallah:
This should be the right, because at least the doctor now will get the second chance to be examined that opinion and reexamined their diagnosis so that is my first way of saying the first wave of AI will be the AI assistant way, where, no, the doctors will not be replaced, the doctors will be supplemented by the AI to significantly reduce the probability of a doctor making a diagnosis mistake.

Amr Awadallah:
Same thing with driving. If you look at the first wave of autonomous driving, if you put Waymo aside to look at Tesla, for example, I have a Tesla. Tesla has the autonomous driving and it’s fully full self-driving, but it still requires you to be holding the wheel. and I think that’s because what what’s happening? You’re saying, oh, are we going to let that system do most of the work?

Amr Awadallah:
But you still need to be paying attention for that. If something really crazy is happening, we want you to engage because they might not be able to do it. I think the reason why they ask you to hold the wheel, those for liability reasons, is they want to say if the accident happens, those of you who mistake your hands on the wheel.

Amr Awadallah:
And that’s the bigger problem here. So if you ask me about this, if you go back to this question about doctors and will AI from medicine be all over the world? I think it will be, but it might be in the US at the end, not at the beginning, because the US is such a defeatist, country where you have liability rules and you have lawsuits happening for anything, even the smallest, most ridiculous things.

Amr Awadallah:
especially in medicine. And that’s why, by the way, the health system is very expensive and very crazy here. But in other countries, like if I go back to my home country, Egypt, where 00 million people don’t have access to good doctors, they just call up their aunt and they ask their aunt or auntie, I have a headache right now.

Amr Awadallah:
My stomach, what should I do? And their aunt doesn’t go to the pharmacy and get this medicine. And that can be a really bad medicine. And that happens every year. That’s really, really bad. And I don’t know if it happens in India and Pakistan and other countries like like Egypt. So. So imagine now instead of the ENT, you have this very, very strong, knowledgeable NLM on your phone that can answer the medical question way better than you.

Amr Awadallah:
And it’s not as good as a doctor, but it’s way better than. And isn’t that a major improvement for developing countries? I think it is. And I think it will really kick off there first, by the way. Like we’d see that happening. It’s my prediction.

Raja Iqbal:
So the way then I, I look at your answer as it is, at least for the next few years, or the foreseeable future, it is going to be more of a decision support system as opposed to a decision making system in most especially the ones that are regulated. And I also get the sense from your answer that.

Raja Iqbal:
Well, and, you mentioned, Egypt or Pakistan or India. if the regulatory requirements are not that stringent, then, overall it will improve the access to healthcare for people, right? So yes, I.

Amr Awadallah:
Mean, I an education education is the other huge thing. There is many studies that show you that if you have a dedicated one on one teacher, if you are growing up as a kid and you have a dedicated one on one teacher working with you, that your intellect and your IQ and your ability to grow in life will be two standard deviations higher than if you were in a class with the other students, right?

Amr Awadallah:
That’s the same you. That’s a control. There was a compulsory the same you. That’s not like a function of oh, the other kid was a porter or richer. No, this is the same you if you were in this environment versus that environment. So imagine now with AI we have teachers that are the best teachers in the world. So they study after they model after the best teacher you can find for physics, the best teacher you can find for math.

Amr Awadallah:
And now all of the kids are learning from them on their phones and with the voice, with with a proper figure looking at them and looking at their homework as they’re solving it with their hand at correcting them. No, do it this way instead, do it that way. Oh, you made a mistake. Here, double check your you. We’re going to have way smarter kids again from the developing countries.

Amr Awadallah:
There’s going to be way smarter kids coming out than ever before because of this technology.

Raja Iqbal:
So perhaps I mean, maybe, as you mentioned, with smarter kids, what happens to social skills then. Right? So because they’re glued to their laptops and their iPads, so what happens? I mean, just out of curiosity, what do you think about that?

Amr Awadallah:
So that’s an excellent question. I love that question. So I would say, two things. First, what makes us superior as humans is tools is both that if you look at, like, the thing that distinguishes humans from any other animal on this planet, and it’s the reason why we are so dominant and we are the leaders of this planet is because we can invent tools and then we can use these tools very, very efficiently.

Amr Awadallah:
Right. So there’s a there’s a very famous study that studied what are the most efficient, animal in terms of traveling long distances as a function of how much calories they have consumed? and the number one animal by far is the hawk, because they fly and they can glide. We can fly extremely long distances on, on a very small amount of calories, and they are by far the most superior animal when it comes to that.

Amr Awadallah:
Now take a human and give them a bicycle, and they beat the hawk. right. So the human with a bicycle will travel way longer distances on a per calorie basis compared to what the hawk is able to travel. And that’s because we are so good at using. So now we have the bike. Do we stop walking? did we not learn how to walk when we were really young kids?

Amr Awadallah:
No, we still don’t have to walk, even though we had the tool available for our disposal because we knew that we need to be doing both of these things, and we need to know when the tool is the right answer versus not the right answer. same thing with calculators. why do we computers have been around for a very long time, but we still need to teach kids at school how to add, subtract, multiply, and divide.

Amr Awadallah:
If you don’t, when you grow up, you never do that. You just use a computer to do that. So why do you still learn the fundamentals? Because we know it’s important. It’s a fundamental skill that we have to learn before we can use the tool to do it. And the same thing applies to social skills, right. So it’s on us.

Amr Awadallah:
Absolutely. We have to be teaching our kids, at the schools. And the same way that teaching a science and they have to be teaching us social skills and teaching us when it’s appropriate to be online and when it’s not appropriate to be online, otherwise it will get out of hand. And frankly, that happened with the social media revolution, unfortunately, is we don’t have we didn’t have proper controls at the beginning, and now it’s becoming an addiction for many, many kids.

Amr Awadallah:
And that’s concerning. And we need to start working on that.

Raja Iqbal:
And that’s that’s.

Amr Awadallah:
Up to us. That’s up to us. It was like, Roger, I want I want to highlight that like we need to, as a civilization, we need to continue to teach ourselves and our kids the fundamentals. And the focus should be on the fundamentals and then immediately on curiosity, meaning building in us and in our brains, the curiosity to be continuously learning and continuously, open to adapt, to do new things.

Amr Awadallah:
And that’s going to be even more important for our kids because we had the luxury. Once I learn data science, I can be a data scientist for the rest of my life. Yeah, our kids are not going to have that luxury that they’re going to learn something. And then I at some point will take it over completely. Oh, now I have to learn something else.

Amr Awadallah:
you won’t be able to do the same career for 00, 00 years. You will have to evolve and learn so that the the key skills that I and a, try to, imbue in my kids and I advise everybody else to be doing the same is number one curiosity, have them ask why questions all the time, and you ask them why questions all the time.

Amr Awadallah:
Get them to be thinking all the time. That’s number one. And number two is adaptability. To have that the willingness to change what they’re doing. Yeah. Unfortunately humans most of us were born with the fear of changing. Right. Once we find something, we love it. And we also do it. We don’t go learn the new thing. Know that that has to.

Amr Awadallah:
You have to tell your kids, know that we cannot exist. If you if you have that fear, you’re not going to succeed in life. And then last but not least is to have resilience. is very, very important. You need you’re going to lose your job, find and withdraw those. You don’t find another job. You’re gonna fall down.

Amr Awadallah:
Stand up again. Pull down. Stand up again. You got to fail. Succeed. Feels like you need to have that resilience is it’s not always going to be good. There’s going to be ups and downs all the time. Through your career.

Raja Iqbal:
So like, teach them to be an entrepreneur, throughout their lives. Right. So feel and then try again and succeed and fail until you succeed.

Amr Awadallah:
Like that. I think that’s very well summarized. Rather like I think the key skill of the future is for all of our kids to be thinking like owners who think like owners and managers that control their life and what they’re doing. And by the way, they’re going to have tools that allow them to do that. Like our kids went, oh, I have a great idea.

Amr Awadallah:
I’m going to make a business around the computer, create the app, create the marketing campaign, create the pitch deck for the investors and maybe the investor that I would pitch to invest the money. Like it’s going to be a different world that they to be living in. And absolutely, I would say that’s a very good summary. Teach them to be an entrepreneur.

Amr Awadallah:
Yeah, I love that.

Raja Iqbal:
Yeah. So going back to, you know, that, maybe the technical or business side of, Victor and the, similar businesses. So do you think that, applications like Victor, and perplexity and so GPT and there are many, like this, do you do you think that this is going to be the future of enterprise in web search?

Raja Iqbal:
so there is this, discussion that often happens that Google, is in trouble because, now perplexity is there. And, I have my own opinion on that, but, I don’t hold off, expressing my opinion. I would like to you.

Amr Awadallah:
Would I would love to hear your opinion. So maybe I’ll express my personal opinion.

Raja Iqbal:
It’s very it’s, for instance, for perplexity. Right. So I think, supporting the infrastructure, first of all, at that scale, a free service like this, I know how the pricing works, but, these kind of things, so, you know, having something like Google, it would be a very interesting business model that they would then have to build.

Raja Iqbal:
I mean, will this own little ads, I don’t know. And then going back to the hallucination problem, then my opinion is that, Genii, as we see right now, we are far from 00 at the moment. Right? So we’re very far from that, right? I see that as a something that is very vertical or task specific.

Raja Iqbal:
And the given the amount of data that, you have, perplexity me have access to. Right. So it may not be as easy, that is to give a very, a detailed answer on a specific topic. Right. So I think that hallucination problem partly it is the correctness problem because, the scope is so broad.

Raja Iqbal:
It’s a general purpose search engine. It’s not. So if it was perplexity for a very specific legal type of documents, I think I would be okay. Perplexity on quality. I would be okay. Okay. But if it is all knowing perplexity, it is still there is going to be some perplexity. Maybe that’s how they need it, right. So and then, of course, I mean supporting the infrastructure like this.

Raja Iqbal:
in the longer term they will need, a lot of money actually, to usually survive. So that’s my big I would like to hear from you now.

Amr Awadallah:
So I agree with you, but I’ll add something on top of that. I think they can do it. Like from a technology point of view I think it’s doable. Like it’s like technology is not the problem. Money is the problem. As you correctly highlighted. That would require a lot of money. The business model is also a big problem is like, how are we going to fund this?

Amr Awadallah:
as we grow? But the bigger problem is, the brand value of Google. You cannot underestimate how strong Google Brands is. I competed with Google Search, Yahoo for a number of years. I was partners with them in the early years, competing with them in the three years. So they are a very, very formidable competitor in terms of the markets, the presence that they have, the Chrome browser that gives them lots of dominance.

Amr Awadallah:
they are the default in, on iPhone devices or Android devices like that. Dominance is is just you cannot underestimate that. Like penetrating that is it is really hard. And maybe that’s why the US government a couple of weeks ago, is saying now that this is clearly a monopoly and needs to be, broken down and hand in different way.

Amr Awadallah:
So that is the reason why it’s very hard to succeed against Google. They just have a very, very strong brand. And then they have a very, very strong business model that’s already feeding this beast. Right? Like if you’re starting from scratch, you don’t have that. So Yahoo tried very hard. We couldn’t do that. Yeah Microsoft tried even hard.

Amr Awadallah:
But Microsoft spent hundreds of millions of dollars over maybe billions of dollars. And they only have 3% market share. Right. And by the way, they have the same thing that perplexity has right now. Right. If you go to Bing, Bing will give you the summary with citations and everything that you get from perplexity, but maybe short term, not as longer.

Amr Awadallah:
I like actually perplexes longer mode. there’s another company called you that come in the space as well that you might be familiar with. So I really wish them the best. I wish them that they would succeed. But at the same time, I know firsthand it’s extremely hard, extremely hard to win consumers over work. Google extremely hard.

Amr Awadallah:
Right. So that’s going to be the challenge, how much money you need to spend to get the consumers. It’s not how much money spent to build the models. It’s going to be how much money you need to spend on the business relationships. How much money does Google spend to give to Apple? I think to be the default on public will never be able to pay that ever.

Amr Awadallah:
Right? So we’ll see if that gets there. So I, I want to I want to take a step back and, highlight something important to you. Rosa is, sorry. Do you want to add something on perplexity. Yeah.

Raja Iqbal:
I think it is also the ecosystem part. Right. So, you know, yeah, Microsoft tried to beat, Apple and, they brought some of the Nokia devices there. They were one of the most amazing pieces of hardware, but they didn’t have the ecosystem surrounding it. Right. Similarly, Google has been trying, you know, Android ecosystem.

Amr Awadallah:
And Android success. I mean, I think, Google’s try with Android is amazing. Now, Google Pixel, that own line of phones, of course, that’s still weakened, but I love the pixel. I have a pixel that’s well, I use the, I think I think it has one of the best cameras out there, but, but yes, it’s it’s very hard for Google’s Pixel to really stand out to Apple.

Amr Awadallah:
Like Apple, it’s just amazing. Samsung itself also is amazing as a company that makes phones so. So this is the opening system and that is the hardware. And the hardware is a different story. Like it’s hard to compete.

Raja Iqbal:
And the ecosystem that they the way Apple is positioned right now in the, in the smartphone market, similarly, Google is positioned in the manner and the search market, right? So I mean, so, you know, I was trying to bring that analogy that it is not just the product that matters, it is the ecosystem that is built around that product.

Raja Iqbal:
Right? You know, people don’t only use a single product. I mean, so the product is the ecosystem, really. Let me put it. or rather, the ecosystem is the product.

Amr Awadallah:
you’re absolutely right. You’re absolutely right. And that’s a big part of that gives these companies at the beginning advantage. But at the end, when they abuse these systems, gives them a monopoly. Right. So, so in the same way that the US government went after Google, I’m pretty sure they’re going to go after Apple for the same thing.

Amr Awadallah:
They need to open up their ecosystem a bit. They cannot have iMessage be on the iMessage. They cannot have the contact exchange we all need between iPhones. And you cannot do it between any other like they were. There is some things they’re doing or they had that was the charger they had that you have to buy the charger from Apple eventually.

Raja Iqbal:
Right. So you know now you I think you are coming with the usb-C.

Amr Awadallah:
But we have to do that. We have to force them to do that. Like the, the European Union have to go to them. You must do it right. Because they don’t want it. So that because they now have the control and they control the ecosystem. So that’s, I agree with you 000% like that. There needs to be some more changes in the ecosystem side to be able to attack other ones, to be able to compete with Google and compete with Apple.

Amr Awadallah:
But going back to, your the premise of your question first unanimous in the same category as perplexity and GPT sir, I want to say, yes, we are we are retrieval of my generation systems. But our focus at, victory was just to be very clear. It’s business to business, right? We we provide our technologies to businesses so they can build these applications within their organizations or for their consumers as opposed to complexity.

Amr Awadallah:
Was building this for you directly meaning for consumers not click or search for GPT, which is also going to be directly for, for consumers. So that was one highlight that, the key caveat, now, is that going to be a future important part of every enterprise? Oh, absolutely. There’s no question every organization will need to have an amazing internal, assistant that can help all of their workers and become a lot more productive.

Amr Awadallah:
Right. enterprise search is broken and everywhere. If you look at enterprise search and even at Google, by the way, Google, internally, we have something called MoMA. I left Google a few years ago, so I don’t want to make lots of, comments about where it is today. But yeah, it was really that like very rarely you will get good answers, but leveraging AI towards that problem and do it in a smart way, taking that connection between people and so on can give you way, way better answers.

Amr Awadallah:
And not just answers can help you do new things. You can say, oh, I’m working on this new presentation for Project Chevron. I mean, they’re speaking on name, but I don’t mean Chevron. The company and helped me create this presentation. They I would go with search would find the best presentations on Chevron. The more recent ones would know from the social graph who was more involved with that topic versus not.

Amr Awadallah:
And with the beginning of that presentation for you. And then you can go quickly. So we need these things. We need these things in our organizations. So if you ask me the question, will this be, an essential part of every business in the future? Yes, it will be, but it’s more than just enterprise search. Rather, I want to highlight that it’s not just about enterprise search.

Amr Awadallah:
This is about software overall. Like so it goes beyond enterprise search. Every piece of software, every app on our devices will have an interactive component to it where I can tell the app what I want. So one of the layman examples I give sometimes is, photo editing software. So imagine you’re out with your family, in Seattle, you’re on the beach, there is a nice sunset happening behind the water, and you take this perfect picture of your family, and you go back home and you’re enjoying the picture.

Amr Awadallah:
But then you notice that one of your kids had their finger up their nose, right? And they did that on purpose just to make the picture about to go and fix that with it. With a mobile app today is very hard, right? You have to go now. If you’re using Photoshop, you have to read the manual, look at YouTube, figure out the steps how to move the hand, correct the shadows within the space properly behind where the hand was the correct.

Amr Awadallah:
The lighting is incredibly hard to do right? The flow, the steps. But it’s doable, but very, very hard to do in the future. Meaning a year from now I’m not looking two years from now like you now you from now you just tell Photoshop, can you please, take my sons, finger out of his nose? And the hand would move away and he’s thinking about the those.

Amr Awadallah:
So this that we’re talking about, this is really why we’re talking about your Roger. This is why I’m excited about this space. Is every single piece of software around this will evolve in a way that will allow us to work with this. Like we work with each other.

Raja Iqbal:
And so in general, then, if we look at the society then. Right? So future of work and society and in general, we see a lot of things that fundamentally, I mean, even for us, with the past, I mean, we have been this space for some time. We have seen things evolved. So in general in society, what is the most exciting application of generative AI that you’re looking forward to?

Raja Iqbal:
It does not have to be in terms of technology, in terms it could be terms of social good, in terms of, you know, the impact. What would that be? If you would like to see, generative AI change, one thing or two things are three things. What would be those things?

Amr Awadallah:
Yeah. So first, I’m biased. I think the AI assistance that we’re building, is that’s and I’ll give you a very quick example of it. So, just to show you that the level of impact it can have and how it can truly change things in significant ways. So one of our customers, it’s called oBike and oBike and is the largest manufacturer in the Middle East, and they have hundreds of thousands of workers in their in the factory.

Amr Awadallah:
And sometimes the machines in the factory break down. So the workers are not engineers and technicians. They don’t know how to fix the machine. They have to call the technician. The technician can take 2 or 3 days to show up, which is downtime for the factory. So what they have done, they took all of the manuals of all of their equipment in their factories.

Amr Awadallah:
They took all of the troubleshooting tickets, with the steps of how these measurement machines were maintained, repair them fixed in the past, uploaded all of that into our system, connected, the prompt API back into the mobile app that the workers have in the factory. And now the workers fix the machines themselves. So when there is a problem in the machine, the worker takes a picture.

Amr Awadallah:
They describe the problem, they get back the steps for how to fix the problem. And then they apply these steps. How is that beneficial? It’s beneficial in so many ways. First, the downtime is reduced. The machine can be repaired so quickly. Second, you don’t have to pay the expensive technicians. Are the workers. Get upskilling. Actually, the workers are learning how to fix the machines themselves right now from because they have this ChatGPT like experience where they’re talking with others, working with them to the steps for how the machine they fix the machine.

Amr Awadallah:
They do that once, they do that twice. And now they became an engineer. or at least for the for that type of equipment they now know how to fix. So how is that going to change? The world is going to change the world around us in every single discipline. We have lawyers doing that. We have, ultrasound, radiologists doing that.

Amr Awadallah:
We have, people doing social work doing that and helping the social workers become more effective. We have people on oil rigs that are worried about, violating a policy that might create an explosion that have the AI system watching them and telling them, no, you should not do this. This is against policy. You should do that.

Amr Awadallah:
It’s going to change everything that we do. It’s going to make us a lot more efficient, a lot more productive, and it’s going to make the average human like this worker in the factory do things that normally they couldn’t do. Now, the other amazing thing about this worker example I just gave is the fact that, the platform that we have is cross lingual, so the language doesn’t matter.

Amr Awadallah:
The neural network, memetic models, everything is cross lingual. So even though the documents is on an English, the workers are able to ask the question in any language they would like. So some of the workers ask in Arabic, some in Punjabi, some in Mandarin, some in French, and they still get the response in their own language, even though the underlying source of knowledge was in a different language.

Amr Awadallah:
And that’s a great benefit for such organizations. So that’s my biased answer, right? As I think we are a big department. Now, if you ask me what is the most impactful AI technology, right? You right now it’s, the cogen technologies, like I can tell you, like our productivity as an engineering team at Pixar is 000 times more than it was ten years ago because of Kodak, right?

Amr Awadallah:
Because of using things like Copilot and GitHub and, so on. It from cloud is amazing. Like if you haven’t used it for coding, you have to be looking at right now, just ridiculously good.

Raja Iqbal:
Better than copilot or GitHub.

Amr Awadallah:
Copilot by far. Yes, but so copilot is good for the smaller. Complete this task. Complete this function. Help me debug this so that for bigger projects, for bigger kind of thing, it’s really will blow your way. Like give it a try. It’s truly, truly, next level like even compared to GPT. So that’s the most impactful today.

Amr Awadallah:
Now if you ask me, what is the most fun today, the most fun today is the image generation, music generation and video generation technologies like. So I just a couple of days ago, if you go look at my Twitter, look at our dollar, you’ll see me. I just post a small video I made. I was dreaming, I made in two minutes and I couldn’t believe I made the video in two minutes I was dreaming, what if there was a big pizza?

Amr Awadallah:
A humongous pizza the size of the Colosseum in the middle of a desert with a great sun rising behind it, and a big crowd of people eating the pizza like surrounding the pizza and what would that look like? And the video came out looking exactly like I was dreaming today with amazing quality of graphics. Amazing. like you couldn’t even tell it’s a fake video except for the pizza.

Amr Awadallah:
The size is huge, so it cannot be real. So so that’s amazing. And I see that as well. As three years from now, any of us will be able to create, the next matrix or the next Star Wars just by ourselves. We won’t need a big crew. We won’t need a hundred million funding to do Avengers and Infinity Stone.

Amr Awadallah:
We’ll just be able to do ourselves. And this amazing movie will come out on the other end. So that that’s very exciting, very, very exciting. But any of us can create any app we would like any movie with. Like any music, we would like a cure. Anybody that has a sickness teach our kids in a much better way because we have these amazing modules on our side.

Amr Awadallah:
I’m so optimistic about the future from that perspective.

Raja Iqbal:
So seven generations from now, people would still be watching Mission Impossible 00 or something like that is going to be the case. And maybe I can I can create my own mission impossible.

Amr Awadallah:
I would say before 00 something iterations is really way deep in time, and that’s all the time. If we go to this mission that we’re living longer and we’re staying longer before we have kids, but we don’t still be alive in seven generations, that’s the bigger question. And who.

Raja Iqbal:
Knows, maybe it must be that being he he doesn’t need so.

Amr Awadallah:
Younger every guy definitely.

Raja Iqbal:
That keeps getting younger. Right. So that’s great. So out of all the startups that, there are many actually that are quite popular, get a lot of press. So besides the vector, what are some of your favorite, startups and why do you find them interesting? Or why do you think that they are likely going to succeed?

Amr Awadallah:
Oh, that’s a very, very good question. So so I see this as different things. Like I can find a startup, an application very interesting, but I could be skeptical of their success. Like perplexity I use perplexity and I love their UI and I love them. But that’s got the other success with.

Raja Iqbal:
Maybe we can separate the two parts. I totally get that. Maybe I’ve done the question correctly. Right. So the idea is wonderful, but there’s some skeptical.

Amr Awadallah:
Yeah, at all. For example, another one I use right now, which I’m very, I love it and I use it. There’s runway ML, runway ML. They do their video animations based on images, but sort of when it comes up might kill them because it looked way, way better. It looks way, way better. And they can do way longer videos.

Amr Awadallah:
So we’ll we’ll see how that plays out. But if you ask me something so you’ll see a company right now that I really respect and I really love what they’re doing. So I have to say I, I really respect meetup, and like the stuff that MIT AI is doing by releasing these amazing models and not just llama, the voice detection model.

Amr Awadallah:
They have so many. The segmentation, the image segmentation, they do segmentation in real time. Video segmentation right now, truly, truly amazing stuff. So they are not just helping themselves. They’re helping the world in an amazing way. So I have to say, I have to give them kudos. And I don’t say that lightly. Like I, I was actually, not a big fan of Facebook historically, but given these recent changes with meta AI racing, all these amazing models, I’m very, very, very, impressed by them and and very positive on them as well.

Amr Awadallah:
I think they will do some amazing things leveraging this big magic. That’s true as well. in terms of smaller startups that I am very impressed by, I would say, very impressed by Mistral and what Mistral is doing. now, what do they succeed as a business? We’ll look at that. They will get some sales. Incredible people.

Amr Awadallah:
One thing that models but open source is hard to monetize. I want to see how they want to do that the longer term. I’ll definitely be very, very impressed by how much they were able to accomplish, given how small their team size is. And let’s see if there’s any other things, than that. So no sono ai, the music generation service.

Amr Awadallah:
I don’t know if you have tried that one. If you haven’t, go get the time.

Raja Iqbal:
I did, but someone was showing, and once again.

Amr Awadallah:
You can go right now and say, you know, let’s make a song about the rasa on Amr, talking about AI, and it will make a song about Rasa and am talking about the AI. It’s all going to come out and bother with music, but in The Thorn in the side, you would like rap, hip hop, classical again.

Amr Awadallah:
Whatever. I just come out, it’s like, it’s really amazing. Very, very cool. Yeah, yeah.

Raja Iqbal:
So how do you keep up with all of that? I mean, I find it as a technologist, I find is, it really mind numbing? Someone will tell me, hey, did you check this out? No, I’m sorry, I I’m not going to be here. So how do you stay up to date? But, what is what is happening in the industry?

Raja Iqbal:
Do you. Yeah, I look at some time reading. Is it some forums, groups, discussions. How do you stay up to date?

Amr Awadallah:
I allocate at least an hour and a half every day to stay up to date. Right. And I stay up to date in two ways. I have a 3 or 4 newsletters that I’m a subscriber to. They think about what are my interests very well. And then articles they send up on all of them. And then Twitter, like Twitter is like following the right people on Twitter and using leveraging the Twitter algorithms.

Amr Awadallah:
Twitter has a feature for top articles which will then all articles from your network. I check that once a day as well, and it shows me all the top articles that my network is sharing and retweeting and so on. And then, historically, that would take longer than an hour and a half. But now or never doing the summarization technology is I can go through a lot more because I summarize most of the articles.

Amr Awadallah:
I don’t read the entire article, I don’t read the entire workbooks. I summarize it. So I have a plug in that I use to summarize that. And then if I find the summary is very intriguing. Oh no. Yeah. This looks like I do. I need to read this one. Then I go deeper and read that one. And I found that increase my, knowledge absorption capacity though it reduced the depth.

Amr Awadallah:
But the depth became higher by about maybe ten, by leveraging these summarization modules. Yeah.

Raja Iqbal:
So you have your own vector, assistant then for all the.

Amr Awadallah:
this is.

Raja Iqbal:
Basically. Right.

Amr Awadallah:
So that’s a good idea. We don’t have that. We don’t we don’t have a plugin that does that. But that would index all of that. So free open source plugin by the way you can grab that will index all of the articles that you’re reading and keep them for you in a rack. And you can go back and say, oh, I saw this article a long time ago about, how people are building a neural networks using, biological neurons.

Amr Awadallah:
And so in semiconductor neurons, can you, summarize that article back again for me? And we’ll just bring that for you right away. So anyway. Yeah, but but now the plugin I’m using is a GPT plugin. It just does summarization. It doesn’t do storage.

Raja Iqbal:
Yeah I see so, what I do is, we have a similar tool, that we have built, but what I do is any paper that I created, yesterday I found ten papers on the AI governance, basically just push them in. And once they are all sorted and organized, I can always go back and ask.

Raja Iqbal:
yeah, I did, summarizing that I found find that is actually a lot of, I mean, call it a force multiplier. It helps me to actually skip that.

Amr Awadallah:
you said at least that is a product that sounds like a good product idea, actually, for researchers and such. Yeah. So you said think about making it available as an open source product.

Raja Iqbal:
Yes, I think so. I’m going to say it was a pleasure talking to you. Thank you so much for your time. I hope you have a great rest of the day.