The Promise of AI: The Founders Perspective

[Music] right gentlemen I'm super excited about this because we've heard from a number of perspectives today and now we're going to be diving into the founders perspective and how they are really using artificial intelligence and AI sit at the center of both of your respective offerings so let's dive into a bit to how you're actually revolutionizing AI adoption at scale and Abdul we'll start with you how are you using Ai and the ultra platform and maybe actually start by what is your Ultra platform okay so um I think at at the center of our uh platform maybe a lot of people here use some of our platforms or one of our platforms at the center of everything we do is Simplicity right we try the whole promise of an ultra platform is for you to be able to conduct your everyday Digital Services in one place using one experience or one user Journey instead of having to go through multiple platforms and multiple user Journeys so at the core of an ultra platform is Simplicity and I think AI help helps really simplify things a lot I think with especially with generative AI we're seeing a massive movement in user experience from click-based Journeys into prompt-based uh Journeys so now all of a sudden it's cool to chat again uh it's not cool to have big UI and a lot of clicking and a lot of Journeys here and there and this is what we do on on bottom through uh through the ultra platform is you're now in a place like Abu Dhabi you're able to pay your bills just by texting you're able to check on your Emirates ID renew your licenses all of that just by shooting a text to the respective entity and being able to redeem that service fa session um moving to you tell me about Builder Ai and how you're helping like non-technical users to deploy gen you know that the AI Technologies look I think it's uh it's important and and really we sat about 20 minutes ago so we're sitting again which is great um you know I'm a big believer that 95% of the world that now needs software um they're not technical but these are folks that have Broadband at home but they're not fiber optics engineers they use satellite TV and they wouldn't know how to put anything into orbit let alone a satellite um yet we're asking them to be really technical product managers understand how software Works in order to now just be able to be in business remember there's a big shift 10 years ago it was a nice to have now it's like this is the minimum needed to be in business and so for us we um have sort of taken an taken a view when we started in 2017 and said the front end experience needs to be common conversational because what can all curators do they can talk and they can order online so if we can give this mixed experience between conversational dialogue um giving options making suggestions making it really visual making sure they feel that they've been understood um we're allowing um everyone to be able to build software and I think you know similar to to to astrotech our belief is these things are a digital utility building software is a digital utility the next 500 million applications are not going to be built like the last 100 million nor are they going to be built by the same kind of people and so um we think the use of AI allows us to be able to go you know to to the largest possible audience and by simply the power of their voice and their imagination they can now build whatever they need I love that and given that you're you know we've spoken today about the the pace of AI and how it's moving so fast you were using AI at the very core of your offerings do you ever feel like you're playing catchup or how are you making sure that you're kind of integrating that Anno of staying ahead is that me you both so I I think with with with technology in general you're always playing catchup right so you'll never be I don't think there is a company or an entity that uh has figured it all out and that's the good thing about technology it has this uh hidden meritocracy of the more effort you put into it new things will come out and then uh you you can you can lead in an industry the minute I feel an organization says we've gotten all figured out they become a Nokia and then they they don't end up very well uh and but again back I think the the the one metric we Chase when we're looking at what technology do we need to adopt because it's it's becoming very noisy right everyone in their dog is like inventing something new now everyone is making a new tool every day and everyone says their tool is the best I think for us it's a very simple measure that we use what do do users actually want like I understand everybody talks about super intelligence and you know uh how can we get two trillion more nodes into a a GPT but U 99% of people like such said don't need two trillion nodes they need that just one node that pays their bill or does their utility and I think the common factor across all successful technology is that it capitalizes on human laziness right I think as we move more and more from the industrial Evolution towards now it companies that are taking more of the backend effort and simplifying the frontend effort to the users you make less effort as a person those are the companies that are deriving more more and more value value as we go so I think for us it's very simple we just keep our focus on what do people actually need what do people actually use and not on how sophisticated as a technology look and I think to sort of to build on that we have a um slightly myopic view of where we actually think we are in AI so everyone seems to think superintelligence is here and jobs are being eaten and the world is being destroyed next week if you imagine where at the day maybe last year was the day and we're a little bit further ahead last year was the day when aola launched the disc and everyone said oh my God the internet's here I was like no The Internet's been here for about 50 years it just got packaged into a dis and so what was the Revelation or the evolution that happened in November it wasn't that large language models were here but was invented 15 years ago transform models and embeddings have been here for eons the big shift was that suddenly you had technology communicating in a channel that was almost entirely reserved for human conversation what does chat gbd look like it looks like WhatsApp or bten and so now you have people that freak out that you have this thing that sits in the same channel that I speak to my mom on second thing whether it was direct or indirect and there's some debate around this the words were slowed down do you remember software if you go to a Google search It's like a instant you go to chat GPT and ask a question and you see the words coming and there's a really important human psychology in this the you remember growing up our teachers used to say slow your response count to three before you answer back it's because human beings trust a response that's slower because it feels like it's more thought out the third evolution actually again nothing to do with technology all to do with design was the vernacular we started using we went from Worlds that were acronyms and Technical terms to hallucination well actually Hallucination is not a confabulation confabulation is a technical term for what happens when it gets the statistics wrong Hallucination is a human emotion and and so I think you know it's really important to sort of see where we are uh I think the second piece for us is being what are you actually using because everything is not a large language problem whe whether a human generates or a GPU generates why did you generate it in the first place today there are 10,000 people in the world approximately building login why are 99,999 people building login why couldn't they use the first log in the first person but in the morning and we laugh right because I'm I'm making I'm sort of being an that's how silly it is to generate everything and you think about how the human mind works I'm sure you have a memory as a kid you're not generating that memory you're retrieving it that's like a Knowledge Graph right if I said to you complete my sentence what am I going to say right that's actually a graph neural network it's predicting my next word on the back of inference and now we want to write a poem about ADF W that's generative in its nature and and we've just assumed and I think you know Abdullah said this we we've just assumed everything is LM and every every man woman in their pet is now building AI but just because because you use a large language model doesn't mean you're building it just you know on that point I think Abdullah mentioned the tapping into human laziness and the thing about memory I canot remember all of my memories as a kid but I feel like my own memory is getting sloppier with the use of AI and Technology because I know that my human brain doesn't have to remember so much anymore how do you kind of see that integrate like is there a point where or what do we use the the intelligence that we've given over to computers what how do humans like look I I think we all have the potential to be superheroes uh the interesting thing here is is the human the cape or is the AI the cape I think everyone's concerned that the human's the cape and actually I'd argue the AI is the cape and and so we're made of organic matter by default we Decay you know there's the tel in our cells get smaller every year they miscommunicate that's why you get sick it's also why you forget things uh and it's perfectly fine that a semantic second brain can tell you what you forgot and how you get back into it it's giving you superpowers I don't see that creating laziness because laziness implies um a a need to not want to do work this is just biological we age and until we solve the Talma problem we will keep aging we may slow it we will still age uh this thing doesn't age because circuit boards don't age and I think that's really the difference um I just to like touch on that I think um there's a lot of this question going around it's not just about laziness but it's just this whole uh thought process that technology should grow and should respect our feelings and should not take away our jobs and should you know keep us healthy and should again you know I I feel like a lot of the time people forget their adults with brains bigger than computers right and we can't really you know the the the most basic human instinct that we were created for with a very long time ago is survival and our survival is not the job of a computer it's our own job it's each and every individual's job and I do feel that there is a big um sense of let's blame all of our employment problems on AI let's blame all of our whatever issues on AI and this happens every time you get a technological advancement people start thinking like the world's going to end and technology is bad and all of that but I always tell people my my very honest message which a lot of people don't want to hear most of the time is your survival in a workplace in your environment is your own worry is your own duty is your own job and it's not the job of technology technology will always outgrow right and if you put yourself in a position where your company has to make a choice between like sat said a computer that doesn't age that doesn't take holidays that doesn't go off on weekends and never fights with its girlfriend and between a human that probably has a lot of different issues that happen in a workplace and the computer is cheaper that company I don't know which company it is but that company is going to choose the computer but unless as a human you evolve and you continue on your survival path and you follow your instinct of you need to evolve as well along with the technology and become a technological operator rather than an executional layer person you're probably interpret you as technology grows and that's that's what I always tell people because because I feel like there's always this sense of technology has to have the responsibility to grow so well and take care of all of us but it doesn't do that and unfortunately sain said it can make us superheroes I think it's always the case that it makes super villains first yeah it doesn't really make superheroes first because the human case for the past I don't know how many years but we've always used technology to kill each other first and then to gamble and make money and then you know whatever is left over we use it for ESG and we take care of the world so I I think for sure it's going to make super villains before it makes superheroes and you know I don't need me to tell you superher super vill are not very nice people and that's and they're not going to be very concerned on Whose job technology is going to take or anything of the sort so again I I I really want to be the guy that came out with this message and I want to be sitting here in 50 years and tell everyone I told you uh uh watch out for yourself because if you're watch if you're waiting for the technology to come and take care of your well-being if you're waiting for the E the ecosystem to grow in a way where it's considerate it has never done that at least in in in in in in the in my information technology has never grown considerably to to be considerate and you know nice to everyone so that's well and I think a lot of it is also because of the beholder that's driving it there's one other angle to look at this you know and it goes to your sort of point of of of laziness or not I think there's two things we have acute scarcity in the world today scarcity of skills scarcity of developers you look at some people doing skilled jobs they don't have enough time companies don't have resources to fulfill the mission that they set out to start with and so the reason I talk about that is the next piece is what is AI in general machine intelligence very good for it's really good for removing variants humans are generally quite variant in our nature so if you have more than one human in the loop in a process you have very high degrees of variance and the second thing is we hate repetitive tasks we just hate them uh and so if you can offload highly variant tasks and offload highly repetitive tasks you're actually creating a balance between in on a global scale between demand and Supply so I don't think anyone has to worry about unless you were doing the repetitive task and you were really comfortable being highly variant no I love that we do have to take the personal responsibility ourselves to develop them as well we only somehow have about a minute left and I know this is maybe not the right time scale to look out when you're looking at AI but where do you see AI on your particular spheres say in the next two to five years is that a reasonable time frame I'm not sure or is that just a lifetime in AI so I think it's going to still evolve massively I think you know if if you were to look at AI today you're looking at probably a six-year-old that's extremely smart I think within the next two years Europe GNA be we're GNA probably hit super intelligence where AI is going to be able to come up with new science Al together like sa said AI now only uses whatever we train it on it cannot create a new science but I think the next two three years we might hear something like AI just created a new form of mathematics that would change everything that we look at and again the only constant is is how Dynamic we are and how adapt to change we we can remain uh as a company we try to simplify everything to its most basic form and serve it to our users you know in in a chat based format and I hope we can continue to do that look and I think I always get very nervous we sort of talk about general intelligence super intelligence because I think the point of the singularity is the point at which we can't imagine so therefore if you took that philosophical argument forward we really can't imagine when that's going to happen either um I think uh you know I I would even say we're even younger than six but I agree with the general deciel I think we're you know just about a totle of walking and then the reason I say is you imagine a child when they're born in the first 3 to six months um they have very limited access to information and they're still building a value model right for example I'm sure you've said to your kids be nice be nice means many things like don't hit your sister don't bite the teacher a number of others but they learn through reinforced learning what that means today what we've got is a child with the world's information and almost no value system and some set of basic rules on top so it shows you just how much Evolution that still needs to come I give you a second example and this is my favorite one but imagine you're talking about American independence and you're trying to teach a child in London versus a child in New York we know that sovereignty is based on culture culture is based on a collection of history history in itself has two principal views at least of what happened so what do you teach the child in New York versus London about American Independence if it's a single llm that's running it because the problem you have today is volume is sort of some sort of authenticity for veracity and it's not correct so you know these aren't technology problems but these are problems that we're sort of getting our minds around and and also show us that just how early this this thing is because you can't just give an answer to that that answer is contextual based on it because you're trying to maintain history other words what you've got is a homogeneous world where there's only one version of the truth we know that's never the case well gentlemen we are out of time that's been such a fascinating discussion I'm sure you'll agree please join me in thanking session thank you so much