I believe AI is making this an especially good time to build a business. On 2024-11-09, I wrote, “We’re living in the best times to build a business.” I shared a post by @tsarnick that attributed to Sam Altman the claim that one person with 10,000 GPUs could make a billion-dollar company; that was a claim in the shared post, not a result I reported myself.
On 2025-01-23, I reacted to a video by saying, “This video was wild” and “What’s coming is amazing.” I expected AI to let me grow my business way bigger than I had thought with a more lean team than I had thought. I did not identify the video or specify the business’s future size.
On 2025-03-04, I said I was diving deeper into AI automations, with the aim of making high-performing teams even more effective while keeping things lean. I did not name particular automations or describe implementation details. This connects with small elite teams beat headcount.
I predicted that AI would eventually be able to do 50%+ of the jobs in a business; I gave no timeframe or evidence for the forecast.
I recommend finding a small, high-performing crew and enabling them with AI rather than hiring many more people. I think this can help profits grow faster and make it possible to share the gains better. I did not specify how to implement that sharing. This approach also relates to systems and discipline prevent bloat and lean stores as cashflow machines.
On 2025-04-12, I recommended hiring an AI before hiring another person. In a reply later that day, I argued that hiring too fast can kill a business because payroll is expensive and many businesses cannot afford it; I said many could benefit from the AI tools available at the time. I did not specify which tools or circumstances I meant. This reinforces my preference for enabling a lean team with AI rather than adding headcount.
On 2025-04-13, I said I have 10+ people on my team, love my crew, and will try to keep things lean whenever possible; I said AI makes that more possible than ever.
On 2025-12-05, I agreed with a post by @iamshackelford and added my own claim: “One person with a Claude subscription and a laptop can outperform an entire agency team.” I gave no evidence or examples for the comparison. The quoted post’s discussion of five agency roles and Meta’s claims about narratives was @iamshackelford’s, not mine.
On March 13, 2026, I said that hiring “cool humans to use AI employees” is better than hiring AI employees directly. I did not define “cool humans” or give evidence or criteria for the comparison; it reflects my preference for people using AI, consistent with my support for lean teams enabled by AI.
On May 16, 2026, I said I thought an interview shared by @davidsenra might be the most valuable interview I had ever seen. The shared post attributed to Applovin founder Adam Foroughi the claim that one great person using AI could be as powerful as 100 people. In the excerpt, Foroughi said his business was expanding but he was not hiring, and the available text cut off as he began explaining why. These were claims and statements in the shared post, not my own hiring decision or a complete explanation from Foroughi.
On 2025-12-05, I claimed that one person with a Claude subscription and a laptop can outperform an entire agency team, though I supplied no measured results.
Related: small elite teams beat headcount, business systems and leverage, lean stores as cashflow machines