Human-First AI Pulse

A 28-year study followed people with childhood ADHD into adulthood to see who ended up starting a business (Tran et al., 2026). For highly intelligent men, ADHD nearly tripled the odds. For women with ADHD, the pattern flipped. They started businesses at close to one percent, the lowest rate of any group in the study, and higher intelligence did not move the number. The researchers are careful here. This is structural, not a question of capability. The same traits that get read as drive in men get read as too much in women. Here is the human read. If you are a woman with ADHD, and building has always felt like it costs you more than it seems to cost everyone else, that feeling is not a character flaw. It is a documented headwind. Naming it does not make it heavier. It moves the weight off your sense of who you are and onto the system that put it there. You are not failing at an even race. You are building inside a rigged one.

Two researchers writing in Harvard Business Review put a name to something a lot of people feel but rarely say out loud (Hinds & Leonardi, 2026). They call it botsitting: the work of supervising, correcting, and cleaning up after AI. The data behind it puts botsitting at roughly 6.4 hours a week, more time than people spend using AI to actually produce anything, and only 13 percent say it has improved how their organization performs. Here is the human read. AI does not delete effort. It relocates it. The work moves from making the thing to checking the thing, and that second job is quieter, harder to see, and every bit as draining. If your tools feel like one more thing to manage instead of a weight off your shoulders, you are not using them wrong. You are feeling the cost this study finally measured.

Two researchers writing in Psychology Today draw a line worth holding onto (Hoang & Nguyen, 2026). AI is very good at prediction. It can surface patterns, generate options, and tell you what usually happens next. What it cannot do is judge: the unusual observation, the personal conviction, the willingness to run the experiment that looks irrational on paper. That part is still yours. Here is the human read. If AI has left you with more information and less clarity, the missing piece is not another tool. It is a decision only you can make. More analysis is not the same as knowing what to do. AI can hand you ten possible futures. Choosing which one is worth building has always been the founder's work, and it always will be.

A leadership piece in Harvard Business Review opens on a CEO whose town hall remarks were written by AI (Lancefield, 2026). They were polished, safe, and hollow, stitched together from old speeches instead of a real answer to the moment. The argument: leaders have to be deliberate about which work stays theirs, because judgment, voice, and presence are the parts AI cannot produce. The human read reaches past the C-suite. Whether you run a company or a business of one, some of your work is not yours to hand off. Your voice is how people know it is actually you. A tool can give you a competent draft. It cannot decide what this moment asks of you, or say it in a way that sounds like a person your people already trust.