GitHub CEO Thomas Domke -- The One-Person, Billion-Dollar Startup
Channel: Alex Kantrowitz
Published at: 2024-07-24
YouTube video id: oU2Qj_CENVE
Source: https://www.youtube.com/watch?v=oU2Qj_CENVE
so the example of the $1 billion value company created by one person do you see that as something that's feasible I think so I think you know the the you know I've seen a bunch of examples where small companies like Instagram comes to mind you know that that started really small and and by the time they got acquired they were still very still very small WhatsApp um is a similar example um and so I think you know uh that that can exist um the question is a single person company how they also manage in uh support and accounting and all the other things that that are outside of creativity and maybe a Copart also helps them with that um you know answering support questions but I think there's you know it's more fun if you have a smaller team uh of of people available yes um so let me let me read you this example that I saw on Reddit of a coder that was talking a little bit about how they've worked with generative AI to build um so they say it's mind-blowing how quick I can move now they're using Sonet 3.5 which is an anthropic model um I'm pretty sure I could Implement copies of the technical parts of the most popular apps in the app St App Store 10 times as fast as I could before large language models uh I still need to make architectural and infrastructure decisions but stuff like programming the functionality is literally 10 times faster right now and this is the process that they use the first thing they do is they think hard about the feature and probably discuss it with Claude the second thing they do is write a basic spec for the feature uh it's just a few sentences and bullet points and also iterate with Claud on the spec and then they are sure to provide Claude with all the relevant context and ask for the implementation the code so basically what they're doing here is brainstorming an app with Claude specking out an app with Claude and then having Claude code it I mean that is that is remarkable is this something that we're going to see be more common I think we're going to see it at a smaller scale um for small projects you can probably um you know get there even without aot lot of computer science Computer Engineering knowledge for larger projects I think the step missing is the architect you know the software engineering expert that that knows which database to pick you know which cloud provider uh how to make the app you know scale from from 10 users to to 10 million users and um the model can kind of help you with that by giving you options right we have all seen that you ask jpd a question an open-ended question it gives you options and it explains to you kind of like how to get there but to navigate then this you know tree of information you still have to have you know subject matter expertise I don't think that goes away but I think you know if you have a well defend task um that you can describe you know to certain level um to to model or to to a whole system like co-pilot you will um you will have that agent if you want to call it like that um do the job job for you um um to to 90% you know of of what you expect you know the flip side of that is if you think about um when you work uh with with other people whether it's on software on other projects it's like how long can you have a person go by themselves when you give them a task until they're going so far off track of what you actually wanted to achieve whether is what you describe to them right like more often than not we need the feedback loop as as humans we we can work an isolation for too long until we're either completely off track or we come back you know uh with a work result that that isn't isn't really uh what what the what the person that or manager or you know or customer expected us to do and and and and I think this is where you know we have the boundaries of these models of the human can do that because ultimately you know the customer can't describe it um or the manager can describe it to the level of degree that you can actually fulfill all the all the the requirements then can do that by themselves as well that's why we believe the human needs to be in the center the human needs to be involved at the step of the rate to make sure that we're not getting into the wrong direction but this is exactly what the person is describing that they're not only asking Claude to write the code but they're dialoguing with the the AI bot about the different spec and the decisions and how to you know set up the components and things like this and then it builds only at the last step you know I called this uh I don't know what it was I called the second brain it's it's kind of like we have an extra uh you know outside of our brain memory uh chip that that gives us all the information that we can store ourselves and even if you know we have a lot of things that we learn uh uh you know in in University in high school and and and even before that that we can't really store um and and we often forget these things and um um and so the AI is helpful um to to retrieve this information again um you just need to know how to ask the right question and and then work with you that's that's why we ultimately called it co-pilot it it it helps you you know to um have fun with the things that you want to work on and it takes you know uh over the boiler plate as we call encoding you know the stuff that that surrounds all the cre creative part of the process right now with Gmail so Gmail allows me to write emails within uh within the Gmail application and will suggest some text for me as I write sometimes accepted sometimes I don't uh but also like I could just go to Claude and ask Claud to write the email so I'm thinking from your in your circumstances like is there a is there a reason why people should be using like the co-pilot within GitHub as opposed to like having this conversation with let's say an anthropic bot and just having that write the code and then dumping it into the um code editor by the way you can also use the AI to summarize the email so um it's basically the the with the AI and the other side summarizes the email with the AI at which point you can ask the question why not just send the prompt to the other person uh and and and save the time on all the uh you know friendliness and and the salutation and and what not that we put into emails because it's or just become proficient at writing concisely but I think as a journalist I you know I know that that part of of the world is not going to um is is more difficult to prti than others sorry go ahead yeah you know and I think to some degree that will happen and so some to some degree we have so much information around us now that um the summary is is good enough um if if you only want to read the headline and the summary and not dive into 10,000 uh bir article because it ultimately means you have more time uh for other things we all you know dealing with limited attention um limited lifetime ultimately and so if that can short some of these things um that means I have more time for other things coming back to your question now why not use the generic chatbot um the the power of Copart is that it lives you know in the work environment of the developer so yeah you can copy and paste everything that you see in front of you into a generic chatboard and um and have it give you an answer but it's much much more powerful to have the chatboard sit within your uh environment um where it knows you know what files are open it knows what you wrote you know before that um it can look at adjacent tabs it can even look you know uh in developer um in the developer world uh at the output as a debug output is what we call that and the console and error messages and those kind of things and so it has much more context available that helps it to answer the question you know within the specific context of the project you're working on you know a very simple example is that it knows you know whether you like your um like your variable names uh with camel case or capitalized and um or whether you write in in German or English we're just looking at at the context of your file right so one question for you so basically you can write the prompt of what you want and the code will be developed on the back end um or will help you devot to develop the code on the back end is there going to come a point where we're not going to need code at all to build because like if you can just prompt then why do we really need to be in the code you might say we we're already at that point to some degree where you know when you ask and go CH GP ask and CH go and ask chat GPT a question uh you get an answer and you must when you ask a question to plot the chart for example or do a math mathematical com um calculation it um it actually generates a python script that then you know plots that data into a chart and it shows you the chart and it still shows you that step where you see between uh the python script but uh you they could as well hide that um and you just see the see the chart output right like in in many ways cat gbt is giving you an answer without you ever having to worry how that was generated and so yes we're going to see computer systems where large language models are just one building block in addition to code or maybe it's multiple language models and image models and you know time series models and whatnot Plus Code combined um to to generate um all the output um that the developer or the user expects um do we still need engineers then to code yeah because I well first of all you know there's billions of lines of code out there that still have to be maintained you know one of the examples I'd like to get is that most banks are still running Cobalt code that's a programming language from uh invented in the late 50s when Eisenhower was a president that runs on mainam I'm talking about like to build new things like is there going to come a time where we're just going to have prompters instead of well most most developers work on an existing code basis so I wouldn't I would push back a little bit on maintenance um we're building on top of an existing World um I think there's developers have always moved up the attraction thata you know we used to build it all ourselves and it came the internet and we used started sharing software components soal open source um uh you know nowadays most applications are sitting on stack of a thousand uh uh components um already and you're building the 10% lay on on top of that now that 10% layer you know might you know at some point be written uh an 80% by AI or replaced by AI but that means you have more time for the remaining 20% on top of that the pile is getting always bigger and the the developers are are still going to have enough work you know CFT out for them in fact I'd say you know AI has created more work for s for developers because now somebody has to build all these AI systems and we're not at all at a point where can just you know have an AI engineer quote unquote do the job of of a real human and like that doesn't exist and um even if it exists it's it works well not demo um but it doesn't doesn't actually do any real work wait so help me Square the fact that AI has created more work for developers yet developers are more efficient well you mentioned you know all the companies with an AI strategy so the all these companies that create an AI strategy including ourselves you know now are not only building GitHub we also building co-pilot and so now we have you know two products to maintain and copilot as an AI system is still a lot of code that we're wearing their in they out um in addition to maintaining and extending uh the platform that GitHub is with with over 100 million developers and if you look at you know all the companies building AI today they uh they car that out either as a new team as a new Focus area in addition to all the things that they were working on before and um you know there's thousands of new startups uh around the world that have made AI um that their investment goal you know the thing that differentiates themselves from from the existing uh companies and as such you know more work has been created for developers um uh working on AI systems okay so you're in a very interesting spot because you're work you're running GitHub and GitHub is part of Microsoft and GitHub with co-pilot might be the perfect company to implement generative AI because it's one of those things where like there's usually a right answer to the question uh there's libraries and libraries of code uh stored on your platform effectively that makes it not easy but more straightforward to train on and when people are writing the large language model can rely on that history to predict or to suggest what the next bit of cod should be so it's almost like the perfectly suit the most perfectly suited discipline to use large language models for is code and in fact if you looked at the discussion of gender and ve recently there's been a lot of discussion of how it hasn't really proved its economic value outside of coding now that's the bull that's the bare case but anyway I'm throwing it out there for a point of discussion the question that I have and lots of people have is is this now something that what you're seeing in your field where it improves the employees Effectiveness by 55% makes them happy or allows them to do more and build more is that generalizable to other fields and if so if you think that the case then why because that's the bet that Microsoft is making right it's not just coding it's everything you know I think we have um forgotten how many things AI already does for for us or you know we are not realizing it um you know the image recognition in my car the street but that is we're talking about generative AI in particular so we can go on for days about how AI has been you know for feed ranking and computer vision fine but the big moment right now is all about generative and generative has been something that GitHub has written with co-pilot at this amazing moment but that's the question is that type of technology in particular transferable elsewhere I think one one scenario that comes to mind um uh that we are already using at GitHub is um support um and so um you know if you look at our support system uh uh today you actually find the G up support cow pilot um that tries to help you before you you know submit your ticket to human and we actually see um and it generally aners so it's generally uses the same L langage models to to stay within the scope of your question and we see that the number of um tickets that get solved that way uh is above 50% so we know 50% of those questions that go through the support copilot get solved by support copilot and do not get submitted into to human and so such it makes you know supporting uh our our developers um um our customers um more efficient for us as a company so i' say you know that's that's definitely another scenario where we see the um um efficiency gains for us as a company other similarly you know we have an internal tool called octobot you know like octo cat or or or logo um I he my t-shirt and it it helps you know our folks internally um to to solve it problems and our it team is getting in no more than three hours um per it supporter back through that internal tool by just by you know helping um uh employees you know to solve their own IT issues instead of instead of having to talk to a human first and it's all you know along the same ways which uh along the same lines and which is like generating text you know that helps you to solve the task um uh that you're that you're that you have a problem with whether it's in support an it all that's in coding um to to focus on the things that um that you're really that you're really getting value out right okay I want to talk about what the next set of botels might bring uh but let's take a break before we do that so uh we'll be back right after this to talk about the next set of models and a bunch of other stuff so stay tuned we'll be back right after this and we're back here with GitHub CEO Thomas donke we're talking about everything that AI can do and sort of how the advances uh might help Propel not just coding but everything else forward all right so here's what I'm hearing about the next set of models that are coming and I'm talking about like the GPT fs and the cloud 4S and whatever it might be um that there the one thing that I'm hearing is that they're going to be much much better uh for coding and I'm curious to hear your perspective on how much further there is to go for these models to be able to handle code and what you think even you know these models getting even better at coding might pretend for um you know in terms of it might pretend for what you're seeing with the um developer community today you believe you know one of the things these new models will be able to help with is um we call agentic um abilities um solving multi-step tasks um one classic example in uh in software today um in small and big companies is that you don't have to go for too long until you have Tech debt uh until you have code that is old and needs to be maintained that needs to be updated that needs to be scanned uh for security vulnerabilities and if you look into the backlog of most Engineers today on the one side they have all the Innovation you know all the cool stuff they want to work on and on the other side they have all the M maintenance tasks um you know and one of them is burning down security vulnerabilities um that have you know stacked up over time and so we think you know the next generation of models will be able uh will be helping with burning down these security vulnerabilities um in fact you know we already have um a feature and Market that we call autofix um that uh helps with known issues um uh uh and burns those down um but it only only works right now in a single file and as you have more powerful models you can do that in across multiple files basically solving the issue not just in one place but in multiple places and how do you teach like a model to be able to do that like I you know obviously reasoning agentic stuff has all been talked about is breaking stuff down to its component parts and then learning how to work on it one by one I mean it seems sort of antithetical to the way that LMS work today they not antithetical but very different which is that they just kind of like take a prompt and then just spit back a bunch of information you know you described it yourself earlier which is like this multistep is that in the first step you get an answer from the model that you know describes the um the solution and then what you typically do when you reason with the model is you ask it you know about more about the first step and then about the second step and so you're drilling yourself down into this tree of of different steps and um I think as we you know move forward agents will be able to to do that uh to a certain degree themselves and um and the tricky part is then to figure out when do I have to come back to the human and ask the question uh that I need the pilot to make the decision and and not um have the co-pilot basically go down the wrong path okay this is sort of a controversial question in the AI world but I'm going to ask it to you do you think we're going to get to the point where we're talking right now about effectively AI taking the Reign and starting to to build and coming back to the human do you think we're going to get to a point where AI is going to just improve itself it seems like a bit of a trick question I'd say you know obviously we have seen the alpha go and and Alpha fold that already happening B AI has improved itself right and has learned to play goal goal and and then got as good as the best go players and in fact got better and what we also saw then afterwards is that the best go players figured out how to beat the model um and then so even though there was like a period of time when everybody was kind of like depressed that now that the game is wounded uh the best players figured out they can still beat um after for go and so I think you know there's definitely going to be um problems that AI will be able to solve for us as I mentioned you know burning down security bilities as an example and I think most folks will be very happy about that because then gives them more time to work on stuff that they actually want to work on instead of doing the same security vulnerability fix over and over again in in multiple files um you know is the AI you know going to get to the singularity I I don't know um and I think you know if I if I know the answer and how to get there I probably you know I build that company myself you build it okay but um you know I'm joking a little bit but like I think you know we will see if um we'll see over the next few years um if AI can not only do what it is instructed by humans but can actually get to a place where it can create itself in in the sense of you know not getting an instruction first and and kind of like you know produce ideas um today you know while it may appear that a CLA or chbt is gener generating stuff at the end of the day it's just predicting the next word right the next word after that it it has no consciousness because it cannot say no to you it can only predict an answer that says I I don't want to prct the answer but you know it still gives you an output it can be silent if a you Bel right and look it wasn't a trick question like the question is not can the model uh sort of learn to get better as it goes right which is still following the model like the question is like the basic design of these AI programs can AI learn to make them even better like can AI be able to take a GPT 4 and turn it into GPT 5 right that's the real question you think that's going to happen it's I think it's like predicting the future and I I don't know if the if I could if I can I don't know um I haven't seen uh you know I haven't seen any indication that that's possible today um but you know maybe I'm on your podcast again three years you tell me see isie maybe the AI is going to do podcast between the two of us no but like look you know all if you look at the technology today it's it's super powerful it helps uh developers and and support agents and it employees and to to achieve their job faster which ultimately means you know they have more time for other things in life I think that's remarkable and I would I'm not too worried about you know AI taking over these jobs and and replacing them with a fully automated um employee now another thing that I find interesting is this sort of constraint on Computing and I was looking at your Twitter and I saw that you recently praised the fundraising of a company called etched which has built chips that are purpose built for inference which is effectively running these AI models which are extremely expensive to run now but it's much cheaper to run on an edge chips how uh important do you think Hardware Innovation is for these Technologies to be able to be cost effective and grow to the point that sort of the industry is betting on and and then on that note what do you think about Ed it's incredibly exciting that we have silicon companies in Silicon valy again I think that's number one there's Innovation and silicon um and um there's not only Edge there's you know a bunch of companies that are going in the same direction and um I think it's just fascinating to see after you know we believe that um Mo's law is over and and there's no more innovation in chips and we're back you know to a world where there's Innovation across the whole stack um we talked a lot about models and we talked about copilot and we talked about agents which you know is going up the stack but there's also innovation going down the stack you know from from the model to the data center um all the way down to the Chip And um I think you know we are going to see uh much more on that uh in the in the coming years um the cost to run influence will come down with the specialized chips the models itself become more efficient um you know GPT 40 mini uh was you know announced last week um uh which is but much faster and much more efficient and I think you know we are going to have innovation on the top end where bigger models come out and do more stuff and we going to have Innovation on the efficiency side where the functionality that you know a few years ago required you know more gpus and more time is now is now done you know much faster and and I think if you the it's incredibly important to have that because the faster you get the response the faster you're able to iterate whether you do that manually you know why they asking questions or whether they doing that automatically in a in a copile where you need to do multiple steps to to um uh generate code or um you know in copile you're always generating um 10 responses you can actually see them in your editor if you if you open the side panel so because we then want to pick the best one uh for the context you're working on you can cycle through those right so if you can get those 10 faster you actually probably get the higher acceptance rate on the developer because they saw the suggestion uh before they kept typing um whatever they were typing yeah and you just mentioned that opening eye has reduced the cost to use uh GPT 40 I got a question uh I asked like what should I ask you and Alex wilham from TechCrunch he he asked me he's like um why is GitHub co-pilot so cheap when the perceived value is so high why not add a zero what do you think we really happy about the price point I think there's a balance and with every price point and every new product to find between um Mass adoption and and the value getting out of the out of the product um um you know $10 for individuals and and $19 uh for employees in a company uh per month um is a great price to uh for all these productivity gains uh it has allowed us to um uh go to 1.8 million paid seeds and um we're really happy about about um you know the competitiveness of of that price point are we going to get to a point where most of the code that's being generated is generated by Ai and the developers are basically Auditors of that code I believe so yeah I said actually two years ago uh at a conference that uh my Poli make then 80% of code is going to be written by AI um in 5 years so I guess I have three years to go um to to for that to become true um last year we already said that on average 46% of code is written by copilot in those files that's enabled and for some languages over 60% again I don't think that's a bad thing I think it's a great thing because it means uh developers have more time to write the thing that actually matters um the thing that is creative that's the thing that's new the thing is differentiated um and they don't have to write all the boil up dat anymore okay okay and then all right last question for you you said that you have 100 million users on GitHub today you think that you're going to get to a billion with this so I'm curious like why you think AI is going to drive so many people to start coding and then what does that mean for a broader economy I believe that today the biggest adoption blocker is the complexity of the technology the complexity of a language that is not the language that we learn and use every single day when we communicate and programming languages are great because they're deterministic um you know the same thing does the same uh has the same output every time you write it um but it's hard to learn uh it's hard to learn uh when you're a kid um uh it's much harder to learn than playing an instrument um or drawing an image um because you have to learn the thing first before you can produce anything and then you still have to develop your craft and and do it over and over again uh to actually get good at it and I think AI is going to accelerate that massively and you know one 1 billion developers um by 2030 or so is a little bit under 10% of the population depending on where Ro population is going um that's actually a low number if you think about it because um we all use computers every single day yet we are not able to create the thing most people are not able to create the thing that runs on those computers um and I think you know most people are able to go to Home Depot and and buy a screwdriver and and put a school in a wall and I think that's it's it's just a going to be a fundamental skill of humans um to to to be able to control the computer and um create something on them rather then they use that and become a professional software developer that makes money by doing so that's a very different question in the same way that not everybody that has you know some uh uh skills at home on Home Improvement uh is becoming a professional contractor professional um uh musician a professional artist right like those things are decoupled and I think for economy it means that we have a much higher um literacy in uh in computer engineering in computer science in software ultimately and that means we will be able to solve more problems because ultimately we TR strongly believe at GitHub that most human progress is going to be achieved with the help of software and um without that software without software developer we're not going to you know climb the the evolution ladder so people used to say Lear to code to people who lost their jobs first as a helpful suggestion and then as kind of an insult and then they started to wonder maybe they shouldn't be learning the code because that's going to be taken over by AI but your your stance on this is no we're still going to need the coders and we still have code look look the AI and the um you know co-pilot is not going to replace the code the code is just lower in the abstraction level in the same way that you know your your chip in your computer still has an instruction set you know used to do Punch Cards and then we had Assembly Language now they go into very technical stuff but you know the the chip at the end of the day is still um you know lots of little switches um that switch between zeros and ones that doesn't go away it just moves into into a layer where it doesn't bother us as much and it doesn't you know keep us from building the things we we want to build and I think that's the the true power of generi very cool well I think you should release this uh app this flight tracker app that you uh worked on I would definitely like to look uh you know it it looks horrible um it it solves one purpose I know every time you know two years today I I know whether I've been been on that specific plane you know it has every plane has the tail tail numbers like a license plate and so it can kind of track oh I've been on this flight on that exact plane before um but it looks horrible it's kind like asking me uh to uh to go on stage uh uh with Taylor Swift and sing and do it with her I wouldn't I wouldn't do that either even though I sing in the shower right and that's I think that can I well describes like the intention here is like one thing is the freedom of being creative and the other is being so good that you can can become a professional right well toas look you're right the the gates of nerditude have swung wide open and I'm totally into it thanks so much for joining great to see you thank you so much for super fun awesome all right everybody thank you so much for listening we'll be back on Friday breaking down the news with Ronan Roy and we'll see you next time on big technology podcast