The Future of AI Agents with Andrew Ng | Interrupt 26

It's good to be back. This has always been such a cool gathering that you put together.

Yeah.

So I think the hype has exceeded my expectations. And then also the doomsaying narratives also got more traction, unfortunately, than I would have guessed, including the job apocalypse, which I don't think is going to be a thing. And on the more positive side, coding agents probably took off fast, actually took off faster than I would have guessed, and the frontier coding agents. I know the hype is that everything AI changes every few months or whatever, and that hype isn't entirely true, but that does feel true for coding agents, where it feels like the frontier of what we can do with coding agents, and it's still competitive, right? I think six months ago I was almost all Claude Code. These days I'm still a lot of Claude Code, but also increasingly open AI coding agents with a mix of Gemini CLI, and I want to support OpenCode as well because, you know, it's open source. So the mix of coding agents we use has been changing rapidly. I wouldn't have guessed a year ago that I'd be coding so much on my phone as well, right? So these workflows, and, you know, like many of you, I have a Mac mini in my office, so all these workflows changing really rapidly. And then agentic workflows are really starting to make their way into enterprises, that feels pretty good as well.

So about a year ago I was writing about the product management bottleneck, which is this observation that if building becomes much faster, then deciding what to build, or the product management work of scoping the project, maybe getting customer feedback, deciding what to build, becomes the bottleneck. It feels like over the last year the product management bottleneck has become much worse, in a good way, because software building has become much faster. But it turns out when writing software becomes 10 or 100 times faster, not only is there a product management bottleneck, pretty much everything else becomes a bottleneck. So some of my teams have seen the marketing bottleneck, because we can build so many features, our marketers are scrambling to figure out what on earth the engineers did in order to figure out how to communicate it. Previously, if you had a product that needed legal compliance, if you took, you know, three months to build it, waiting a week for legal to sign off is maybe okay. But now if you're building in a day, then wait a week for legal to sign off, that's a legal compliance bottleneck. And this is a design bottleneck and so on. So I've been thinking a lot about how software engineering teams in the future will be organized, and I don't think I know the answer, but increasingly I find myself setting up very small parts, you know, anywhere from one to 10 engineers in a team of often generalists, high-context, highly empowered generalists. They're given a set of very wide guardrails within which they can just run like crazy and build and ship code and even drive decisions like writing marketing copy, you know, that are traditionally outside the purview of engineering. Maybe by definition, let's say you have a team that needs software engineering, product management, a little bit of, you know, you need a new terms of service, you need some marketing copy, you need some design, say you need five functions represented in a team of two humans, then by definition, or, you know, by the pigeonhole principle, right, these two humans have to play more than a single role per human. But the good news is it turns out I don't think I'm a very good marketer, but when I use, you know, AI, right, I'm still not a good marketer, but I'm slightly less bad compared to if I had to do it without AI assistance. So I find these small teams of high-context generalists that are deeply technical, but able to use frontier technology to do a little bit of other roles as well. Like every engineer, you know, frankly, uses AI to take the first draft of a terms of service to then take to a lawyer for the lawyer to finally polish it before it goes out. But I find that these processes allow teams to move much faster. So that's been really exciting.

Yeah, so a lot of the people I work with most closely have deep engineering expertise. And then also are highly empowered, slight generalists that step into other roles and acquire this mix of skills, often with AI supporting where, you know, they didn't have that training. I think people can succeed from any direction. I've seen, you know, product managers become much stronger at coding and then participate in these teams. I think just because so much of AI coding, and engineers maybe had a natural advantage understanding the frontier tech. So I see engineers play this role successfully the most often, but there's definitely a smaller number of product managers. I've seen marketers learn to code in pretty effective ways. I've seen operations people start to build more and more products. I think it's actually possible for people of, you know, any background to learn to do a lot of this, but the largest number of people doing this well right now seems to be people that came from an engineering background. But really encourage everyone from whatever background to see if you can play these roles.

Yeah. So there's one way I've been thinking about the future of software engineering. This is a mental model I have in my head, which is that because of a lot of providers of many tools like, you know, RAG, agentic frameworks, I don't know, things like evals, guardrails, it turns out that these are AI building blocks, they're also non-AI building blocks, things like user interface components or identity auth mechanisms and, you know, front end, back end, persistence, databases and so on. So I think that in computer science we've always had a wonderful set of building blocks, and with agentic coding, building blocks are proliferating because more and more people are building open source or proprietary API-based or whatever types of building blocks. So they're just wonderful building blocks all around us. And so I find that developers that have a good mastery of enough building blocks can often put them together in combinatorially many ways to very rapidly build software. Right? Maybe if you picture a building with Lego bricks, if all I have is like a white-colored Lego brick, you know, I can build some stuff, but it's not that interesting. But I mix in a black Lego brick, a yellow one, and a brown one, and a green one, and throw in some squiggly pieces of Lego, then what I can build grows combinatorially or grows exponentially as a function of the number of Lego bricks I have. And I think of a lot of the building blocks we now have access to as being akin to that too. So I find that the developers that are able to have a good sense of a lot of what these building blocks do, and DeepLearning.AI offers a lot of long-form short courses to help people master these wonderful building blocks provided by tons of people. And then the other challenge is to use coding agents to rapidly assemble these building blocks into the software that you want. One challenge that coding agents have is a lot of building blocks are so new that the coding agents do not know how to use them. Right? I think until recently the leading coding agents, where, you know, Nano Banana was released after the knowledge cutoff dates of a lot of leading models, so they just had no idea how to call the Nano Banana API or even knew that Nano Banana existed. So one project that one of my friends, Rohin Prasad, and I have been passionate about is Context Hub, which is kind of a Stack Overflow for AI agents, for AI agents to get the latest documentation on what are the latest APIs, SDKs, building blocks they can use, as well as a way for agents to give feedback to the documentation to try to improve it for everyone. And so it turns out there are a number of APIs that I personally use that, you know, I find the syntax slightly annoying to remember. But I find that by using Context Hub to load the latest docs, I let the coding agent accurately make all of these API calls for me, right? And so it actually has helped my own coding work, you know, accelerate quite a bit.

Yeah, so we're trying a lot of ways to improve the education experience. I mean, in terms of training, the thing that's clear is that what people have to learn has changed significantly. I think for developers, we have to learn coding agents, learn these building blocks, maybe learn some product management or these types of general skills to make us more effective. So what we learn is changing. And DeepLearning.AI, Coursera, you know, we didn't really try to provide a lot of that training. But separately from what to learn, there's the delivery of training, and it feels like we've been, you know, thinking about how will learning transform a long time, and it feels like it's actually not quite here yet. One thing we launched just several weeks ago is a new website in preview called coderoom.ai, in which the vision was rather than taking an online course, come and have a conversation. So it's not a course, it's a conversation, where the experience we tried to build was for you to come and get on a simulated video call with me, say, where if you feel like leaning back and listening to me talk and present to you in a one-on-one simulated video call about Context Hub and coding agents, you lean back and I just show you some ideas. Or if you want to interrupt me or interrupt my AI and ask a question, you could do so at any time. And one thing I actually personally played around a lot with is replacing videos and slides with JavaScript. And what that means is instead of a video, when I present, when I demo something in a video, if you could, you know, click into the video and type your own prompts or type your own queries into the video window. So instead of a static video that just plays, the video area is interactive, and this creates more ways for people to interact. So check out coderoom.ai if you want and have a conversation with me or with my AI, where I'll present to you in AI form how to use Context Hub and coding agents. But I think we're actually, frankly, still iterating and improving these experiences every day.

Oh, that's live. So imagine if instead of me screen sharing in a video, I am, you know, JavaScript sharing in a video, and so you can interact with whatever I'm sharing because it's running JavaScript code, not a canned video. So yeah, so if any of you play with it, we would love your feedback. But it feels like I think the transformation of education has been overhyped. I think something is coming. But today, taking online courses, you know, I wish we had something way better than online courses. And I will say we have something better than the courses we had 10 years ago. They're much more interactive. It turns out for a lot of our courses, this morning, we actually, this week we just launched a new course on transformers taught by AMD's Sharon Zhou. And I find that rather than just having videos, which is mostly what we had, you know, a decade ago, we now build much more interactive visualizations. There's a lot more fun stuff to play with than courses had a decade ago. But that bigger transformation, I actually spent a lot of time thinking about that.

Yeah. So I think every enterprise, I'm guessing all of yours are excited about AI adoption. One thing we've seen is over the many businesses, turns out one of my teams, AI Aspire, which is an AI advisory firm that my business partner Kirsty Tan and I co-founded, we talk to large businesses all the time, sort of, you know, largest kind of Fortune 50, Fortune 500, G2000 businesses about their AI strategies and transformation. So some consistent themes, right? All of us have invested in the bottom-up innovation, let a thousand flowers bloom strategy, and for the most part, it's not paying off. So CEOs and boards are asking, where is the ROI for AI? I think we should keep on investing in bottom-up innovation. Let's keep on doing that. But it turns out that bottom-up innovation often results in point solutions that drive incremental efficiency gains, which are actually a good thing, but not the transformation, not the broader transformation that AI has been promising us and that I think we should work to deliver. So just to illustrate this as an example, my team's worked with a number of banks. We actually do a lot of work in financial services, but work with a number of banks. And if you think about the process of underwriting a loan, then maybe, you know, five steps: you got to market a loan product, get the application, review and approve the loan, do the final diligence, and execute the loan, right? So let's say five steps of that. A number of teams have noticed that the step in the middle of loan approval, we could use AI to do that. And if we can automate that, then instead of a human spending an hour reviewing the loan application, we could have AI do it with, you know, and that's great. We should absolutely do that. But it turns out that if your entire process of underwriting the loan stays the same except for automating what was previously one hour of human time, that's a small incremental efficiency gain. So what a number of banks have said is, you know what, instead of doing this efficiency gain, which is worthwhile, let's rethink the entire workflow and market a get-approved-in-10-minute loan product, because rather than waiting around for a week for a human to be free for an hour, we can send the loan application to AI right away for a decision. But the challenge with implementing this in a lot of businesses is this takes someone with the broader scope to rethink and redesign the entire workflow, because now you have to market a get-approved-in-10-minutes product. You have to route the application to approval not in a day, but, you know, right away. So marketing data infra needs to be involved. Then yes, AI can make the initial decision, and then final diligence execution process needs to scale up as well. And so I find that bottom-up innovation is really valuable, generates lots of ideas, but often it takes, and that has to be complemented with a top-down motion of having someone with a broader scope to change how all of these steps operate to then create growth. And I find that many businesses talk about cost savings, you know, that's fine, if worth doing, but the more, but I try to push for more imaginative things we could do with AI, which is drive growth, because, yeah, we can only save so much money, but growth has almost no practical ceiling. So the more exciting ideas I find usually relate to driving business growth rather than savings.