Eric Siu breaks down how companies are actually profiting from AI. He pushes engineers toward code-assisting agents like Claude, recommends a CEO-led AI fluency program with four skill levels, and argues companies should build autonomous agents instead of just using ChatGPT as a search tool.
Did you know that 88% of companies using AI right now are not actually making money? So here's what you need to actually do about that. So number one should be easy, but it's not for a lot of companies. Still today, there are a ton of companies, engineers that resist using agent encoders such as Claude Code or using OpenAI's Codex. Any of those are amazing. They help you get a lot more done. We're talking 10 times faster or even faster than that. And any engineer that has used this, they've moved on to bigger things other than just writing code. They're now more code reviewers and they can focus on more strategic things that ultimately help your company move a lot faster. All right, step number 2 is you have to have some type of AI fluency program at your company and it has to come from the top. It has to come from the CEO. Ideally it's a CEO founder because they have that invisible hammer and they can make things happen. So at my company we have 4 levels of proficiency. We have unacceptable, that's level 1. Then you have capable, then you have adaptive, and then you have transformative. Okay, so why does everyone need to be adaptive? Well, the reasoning is because today's capable is tomorrow's unacceptable, and today's adaptive is tomorrow's capable. So you wanna get people just ahead of that curve. And the other thing I would say is you wanna make sure that your team is not just using ChatGPT just to search. You wanna make sure that they're moving to the next level. They're learning how to build agentic workflows. And then after that, they're learning how to use autonomous agents. to do all the work for them. So we like to say here over at my company, Single Grain, where we just don't wanna hire humans to do robot work. And I'll say that again, you don't want to hire humans to do robot work. And so all the things that you were doing on a computer before, or you're used to doing, you know, years and years ago, that's all going away. The whole idea that, you know, you're time constrained, or you don't, you're, there's too much work to be done, or you need to hire more headcount. Jensen Huang from Nvidia has said himself that those ideas are gone. Creativity. Is the only constraint. So if you want more like this, go ahead and comment newsletter, and I'll send you more on AI marketing and business.
