Morgan Stanley AI Leap Report: What It Means for Solo Builders
Prefer to watch? Watch on YouTube · Subscribe for new builds · Read the transcript
In this video I break down a Morgan Stanley report, covered by Fortune, saying a big AI leap is coming in the first half of 2026 and most of the world isn't ready for it. I go through what the report says about compute scaling, power, and jobs, and then what I think it means for non-technical builders, solo founders, and small teams.
I recorded this in March 2026. AI tools and models change fast, so check the current docs and news before acting on any specific detail here.
What I cover in this video
- 0:00 Why a Morgan Stanley report on AI is worth paying attention to
- 0:46 Compute scaling and the "10x compute doubles intelligence" idea
- 1:32 GPT-5.4 Thinking scoring 83% on the GDPval benchmark
- 2:19 The projected US power shortfall
- 4:35 Workforce cuts that are already happening
- 5:20 Why this is a big opportunity for one-to-five-person teams
- 6:51 Recursive self-improving AI as early as next year
- 9:07 What I think you should do right now
Compute scaling and expert-level AI
The main argument in the report is about compute scaling. The amount of computing power going into AI training is growing faster than most people realize. The report points to Elon Musk's claim that using 10x the compute to train a large language model roughly doubles its intelligence, and Morgan Stanley says the scaling laws behind that claim are still holding.
In plain English, the models we use now, Claude, GPT, Gemini, are going to get a lot smarter, because the labs keep putting more compute into training and more compute has kept producing more capable models. If you think Opus 4.6 or another frontier model is good now, expect it to get much better over the next few months.
The report also cites OpenAI's GPT-5.4 Thinking model scoring 83% on GDPval, a benchmark that measures AI on economically valuable tasks. That score puts it above human expert level. So we already have AI doing expert-level work on real economic tasks, and Morgan Stanley expects the curve to get steeper.
The power crisis and the jobs shift
The report predicts two big effects. The first is a power crisis. Morgan Stanley projects a net US power shortfall of 9 to 18 gigawatts through 2028, which is a 12 to 25% deficit in the power needed to run these AI systems. Companies are already turning Bitcoin mining operations into AI data centers, firing up natural gas turbines, and deploying fuel cells. The report also describes a "15-15-15" dynamic: 15-year data center leases at 15% yields, generating $15 per watt in net value. That's why the infrastructure buildout looks like one of the biggest investment opportunities of this generation.
The second effect is jobs. Morgan Stanley says executives are already making large workforce cuts because of AI efficiencies. In the video I talk about Amazon having engineers document their work for eight months, feeding those SOPs to AI, and then cutting 35,000 people, and about Meta laying off 16,000 people to redirect money toward AI. And the trend is speeding up.
Why small teams have the advantage
Here's my take: this is the biggest opportunity for small teams and solo builders we've ever had. In the report, Sam Altman describes entirely new companies built by one to five people that can outcompete large incumbents.
I'm living that. I run my whole content pipeline and YouTube operation with AI agents on a Mac Mini. Each agent has a job: one writes scripts and outlines, one handles X posts, one does intel scouting, and one clips my videos into Shorts. That crew does work that would normally take a team of five to 10 people, and we've gained 5,000 subscribers in a few months. I started vibe coding about a year ago, and now I run 13 agents that do 90% of the work across my businesses. I don't have a computer science degree.
People in the Shipping Skool community are doing the same: one member is building an AI title research tool for real estate, another is automating his entire customer pipeline, and another shipped a SaaS product in three weeks. None of them are Google engineers.
The report also mentions xAI co-founder Jimmy Ba saying recursive self-improvement loops, where AI upgrades its own capabilities without human help, could show up as early as the first half of next year. That is a little scary, and I'll say so. But it's coming whatever we think about it, so the real question is which side of it you'll be on.
What to do right now
- Start building with AI now. You can use Claude Code or Codex to build a SaaS app from scratch, and OpenClaw, which is free, to run agents that automate your business. Even building something small teaches you how the tools work, and learning now is easier than learning later when everything moves faster.
- Figure out what AI can do for your situation. A freelancer might use it to handle twice the clients. A small business might automate customer support with agents. If you have a product idea, AI can help you build and ship it in weeks instead of months.
- Surround yourself with builders. Vibe coding can feel lonely, especially in a small town. Being around people who are also building keeps you accountable and gets you unstuck faster.
Exponential curves sneak up on you. Progress feels gradual, and then suddenly it's everywhere. Morgan Stanley is saying we're at that inflection point now. The tools available today will be much more powerful a year from now, and the people who started early will be in the best position.
Key takeaways
- Morgan Stanley says AI scaling laws are holding: more compute keeps producing more capable models.
- GPT-5.4 Thinking scored 83% on GDPval, which is above human expert level on economically valuable tasks.
- The report projects a 9 to 18 gigawatt US power shortfall through 2028.
- Large companies are already cutting jobs because of AI efficiencies.
- Solo builders and small teams using tools like Claude Code, Codex, and OpenClaw now are best placed to benefit.
FAQ
What did Morgan Stanley say about AI in 2026?
Morgan Stanley's report says a major AI leap is coming in the first half of 2026 because compute scaling laws are holding, and that most of the world isn't ready. It also predicts a US power shortfall and more AI-driven workforce cuts.
How big is the AI power shortfall Morgan Stanley predicts?
The report projects a net US power shortfall of 9 to 18 gigawatts through 2028, which is a 12 to 25% deficit in the power needed to run AI systems.
What did GPT-5.4 score on GDPval?
According to the report, OpenAI's GPT-5.4 Thinking model scored 83% on GDPval, a benchmark for economically valuable tasks. That puts it above human expert level.
How can non-programmers prepare for the AI leap?
Start building now with tools like Claude Code, Codex, or OpenClaw, even if it's something small. Think about what AI can do for your specific work, and spend time around other builders so you stay accountable.
Chapters
- 0:00Morgan Stanley drops the report
- 1:45What compute scaling actually means
- 3:30GPT-5.4 hits expert-level performance
- 5:00The power crisis nobody is talking about
- 6:30Jobs are already disappearing
- 8:00Why this is the biggest opportunity ever for small teams
- 10:00Recursive self-improving AI by 2027?
- 12:00What you should do RIGHT NOW
Transcript
Read the full transcript
0:00Open claw is changing the entire finance sector and Morgan Stanley just wrote a report saying that this massive AI breakthrough is coming right now. And if you are in the finance space, you're probably going to want to pay attention. They're saying most of the world isn't ready for this huge disruption in finance. You know, when Morgan Stanley says something like this, you know, you should be paying attention. This isn't just some random tech blogger making predictions. This is the biggest investment bank on the planet telling their clients to brace for impact. And the timing of this is really interesting because we're already in the first half of 2026 and this isn't some distant future prediction. They're saying it's happening right now. So, I'm going to break down what they're actually saying and why I think it matters so much for builders like you and I. The core of
0:46their argument is about compute scaling. They're saying that the amount of computing power being thrown at AI training is increasing at a rate that most people don't fully understand. And the key insight they reference is from Elon who said that applying 10x the compute to train a large language model effectively doubles its intelligence. Morgan Stanley is saying the scaling laws behind that claim are holding firm. I want to put this into context. We're not talking about some theoretical discussion at some academic conference. This is a bank that manages trillions of dollars in assets telling their investors to prepare. That's a different thing. So, what does this actually mean in plain English for you? It means that the AI models we're using right now claw, GPT, Gemini, pretty much all of
1:32them, they're going to get significantly smarter. Not just a little bit better. Insane. Because the labs are throwing more and more compute at training runs. And the math says that more compute equals more capable models. That relationship has not broken down yet. And if you're using Opus 4.6 or one of these frontier models and you think it's good now, it's going to get infinitely better here in the next few months. And we're already seeing this play out. You know, they point to Open AI's GPT 5.4 thinking model which scored 83% on something called GDP valve benchmark. That's the test designed to measure AI AI on economically viable tasks. And 83% puts it at above human expert level. So, if you think about that for a second, we already have AI performing at expert
2:19level in the real world on economic tasks. Morgan Stanley is saying the curve only gets steeper from there. Now, here's where it gets real for us. They're predicting that this breakthrough is going to cause two major things to happen. The first thing is a power crisis. They're projecting a net US power shortfall of 9 to 18 gigawatts through 2028. That's a 12 to 25% deficit in the power needed to run these AI systems. If you let that sink in, that's kind of crazy. We might not already have enough electricity to power the AI revolution. And that's the bottleneck that nobody is covering or talking about. Companies are already converting Bitcoin mining operations into AI data centers. They're firing up natural gas turbines. They're deploying fuel cells.
3:04The demand for electricity to run AI is about to strain the entire power grid. And there's a wild economic model they describe called the 15 15 15 dynamic. It's basically a 15-year data center leases at 15% yields generating $15 per watt in net value creation. The infrastructure buildout around this AI becoming one of the biggest investment opportunities in this generation. The second thing is jobs are already disappearing. You know that. You see it in the headlines. Real quick, before we keep going, if you're watching this and you want to actually build with some of these tools, not just watch videos about them, then you're going to want to check out our community down below Shipping School. We have a full clawed code course, a full open claw course, and [music] four live bootcamps every single
3:50week where we actually help you get set up from scratch. Like actually set this thing up. Not just watch a tutorial and figure it out by yourself. also provide one-on-one coaching so you could book a call with me. We could share screens and I can help you get clawed code or open claw running on your machine. That's it. No fluff. I built this community because watching YouTube's only gets you so far. We launched it just 3 days ago and we have over 55 members. You need people around you who are actually building, people who hold you accountable, and coaches who can help you when you get stuck. I'll put the link in the description down below. Get in now before the price goes up. And this is part of what I think that hits close to home for most people watching this. Morgan Stanley says executives are
4:35already executing large-scale workforce reductions because of the AI efficiencies. This isn't a future prediction. It's happening right now. We've already seen this play out in real time. Amazon forced its engineers to document work for 8 months. Then they fired them. They basically fed their AI SOPs and then cut 35,000 people. Meta, just this last week, laid off 16,000 people specifically to redirect the money towards AI. And the trend is accelerating, not slowing down. Companies are figuring out that AI can do a lot of the work that used to require large teams. Here's the thing, and this is my take on it. This is actually the biggest opportunity for small teams and solo builders that we've ever seen in a lifetime. Sam Altman said it himself that in this report, he
5:20envisions entirely new companies built by just one or five people that can outcompete large incumbents. Large incumbents are going to fall victim to one to five people teams like you and I. That's not science fiction. That's happening right now. I know because I'm living it. I run my entire content pipeline, my entire YouTube operation with AI agents on a Mac mini sitting right behind me. We've gained 5,000 subscribers in just a few months and all the content is basically on autopilot. The Rizza writes the scripts and outlines, Inspector Deck handles the X post, Ghostface does the intel scouting, Method Man clips my videos into shorts, the Jizza runs the skeri-azit. And I've got an agent crew doing the work that would normally require like a team of five to 10 people. And that's just me.
6:06Inside of Shipping School, the community, we have members building real products with AI. Dustin is building an AI title research tool for real estate. Colin is automating his entire customer pipeline. Faheem shipped a SaaS product in 3 weeks. These are regular people, not engineers at Google. They're using AI to build things that would actually have not been possible 2 years ago. So, when Morgan Stanley says breakthrough is coming and the world isn't ready, I'm not scared. I'm super fired up and I hope you are, too. The people who are already building with AI are going to be in the best position possible when that breakthrough hits. That's one more thing from this report that drove me crazy is they refer they reference Jimmy Ba, who is a co-founder of XAI, Elon Musk's AI company. He's saying that recursive
6:51self-improvement loops could emerge as early as the first half of next year. That's an AI autonomously upgrades its own capabilities. AI making itself smarter without any human intervention. That might sound scary and I'm on I'll just be straight up. It kind of is, but the thing to think about is the question isn't whether it's going to happen. It's going to happen regardless of our opinion. I just want you to be on the right side of it. So, are you going to be the person who spent last year building with AI, understanding how these tools work, developing the skills to direct and manage AI agents? Or are you going to be the person who ignored it and then woke up one day wondering what in the heck's going on in the world? I know which side I want to be on. It took me 35 years to figure out
7:36that key to success is positioning yourself before the wave hits, not after. I worked a minimum wage job for 15 years before I changed my mindset and started building. And then I've been a 6-year entrepreneur in 2020 when I lost my job. I didn't know what I was doing, but I knew I needed to learn some skills and apply them to build a business. You know, the people who are going to win are the ones who start moving before everyone else even sees what's coming. We had family friends over yesterday, you know, five people, adults, and they knew nothing what was happening in the AI space. I don't blame them. You know, we live in a small town, but it kind of should raise a red flag for you that, wow, even though you might be in the AI space, not a lot of people know what's going on. And those are the type of people that don't know what's going on are going to be the ones waking up one
8:22day going, "What happened? How did I just get fired?" You know, that's why I make these videos. And that's why I built the community around AI because I genuinely believe that the next 2 to 3 years are going to reshape the entire economy. And the builders, the people who are actually shipping with stuff AI right now are going to come out on top. So, let me give you a practical example of what this even means. Right now, today, you can use clawed code, codex, to build a SaaS app from scratch. You can use open claw to run AI agents that automate your entire business operation. You can use tools like claw team, which I talked about in the latest video, to have teams of AI agents collaborating on complex projects. A year from now, those tools are going to be dramatically more powerful. And that's the thing about exponential curves and network effects. They sneak up on you. You think the
9:07progress is gradual and then suddenly it's everywhere. Morgan Stanley is basically saying we're at that inflection point right now. So, what should you do? Here is my take. First, you need to start building with AI today, right now. Open claw is free. Even if it's just setting up clawed code or open claw and having it build something small. Get your hands dirty. Understand how these tools work because the learning curve now is way easier than learning from a year from now when everything is moving a thousand times faster. You need to also think about what AI can do for your specific situation. Maybe you're a freelancer and AI can help you make twice the clients. Maybe you're running a small business and AI agents can automate your customer support. Maybe you've got an idea for a product and AI can help you build and ship it in weeks instead of months. Whatever it is, the tools are here right
9:52now and they're only going to get better. You don't need to be a programmer. You don't need to have a computer science degree. I'm proof of that. I started Vibe coding about a year ago now, and I right now run 13 AI agents on a Mac Mini that does 90% of the work for all my businesses. Surround yourself with builders. That's the next thing is, you know, it's a very lonely thing Vibe coding, especially if you live in a small town. Being around other people that know and feel what you're talking about can help you hold you accountable to the things you want to build. That's why we built the community because building with AI can be lonely. It can be confusing when you're doing it by yourself. Well, you when you're in a community of 115 people all doing the same thing, sharing what works, helping each other, debugging problems, doing six live build sessions a week, the
10:37progress is so much faster. I've seen it with my own eyes. People going from zero to shipping a product in two or three days because they had the right support around them. Morgan Stanley's conclusion is pretty stark. They said the coin of the realm is becoming peer intelligence forged by computer and power. The explosion is arriving faster than anyone is prepared for, but you know what? I'm ready for it, and I hope you are, too, because we've been building. I know you've been trying to figure out what to build, and, you know, I've been shipping as much as I can, and I've been trying to learn and apply what I'm learning to these tools and developing these skills while everyone else is just sitting on the sidelines. If you're watching this and you haven't started yet, this is your sign. Don't wait through the Don't wait for the breakthrough. Build now so you
11:22could be ready when this does hit. The people who are going to thrive in this new economy are the ones who started before everyone else figured it out. I got the link to the community down below if you want to join us. We've got build sessions. We got courses in there as well, and you can work with me in a one-on-one setting if you'd like. You know, I'd love to have you, and I'd love to help you take the next step along in your journey. You know, Morgan Stanley says massive AI breakthroughs are imminent. I think we all know that, but to have Morgan Stanley say it, it's different. The world is not ready, but we are. And honestly, that's the best position to be in. While everyone else is scrambling to catch up, we've already been building for the future. Praise God on the timing on this. It's a great time to be alive. We're living in a great time to take your idea and to make it come to life. Never been easier. If you want more video updates like this, make
12:07sure you subscribe. Hit the bell icon cuz we release about four to six videos every single day. We'll see you in the next one. Have a blessed Sunday.
Ready to start building with AI?
Join the Shipping Skool community free. Paid members get live bootcamp calls every week and weekly-updated courses.
Join Shipping SkoolKeep reading
Google Antigravity 2.0 and WebMCP: What I/O 2026 Means for AI Agents
At Google I/O 2026, Google announced Antigravity 2.0 and its CLI, managed agents in the Gemini API, the Antigravity SDK, WebMCP, Chrome DevTools for agents, and Android agent tools. These don't replace OpenClaw or Hermes Agent, but they show Google building infrastructure for AI agents from the browser to Android.
Claude Opus 4.7 Benchmarks: What It Means for OpenClaw Agents
Claude Opus 4.7 beats Opus 4.6 and GPT-5.4 on SWE-bench Verified (87.6%), SWE-bench Pro (64.3%), and MCP-Atlas tool use (77%). It also checks its own work before it answers. The price didn't drop, so I save it for the two or three highest-stakes jobs in an OpenClaw pipeline and leave the rest on cheaper models.
OpenClaw Goes Mainstream: What CNET Coverage Means for AI Builders
OpenClaw just got featured on CNET, marking AI agents' transition from developer tools to mainstream business necessity.
