I got good at 21 devices, then let every one of them go

Hi {{first_name|friend}},

I fell down a fun rabbit hole this week. Someone built a clickable timeline of 87 consumer devices going back to 1976 and you tick off the ones you owned. What I noticed in my own list turned into the whole theme of today's issue and it's a hopeful one relating to AI.

My Take

I ticked 21 of those 87 devices. Looking at my picks, I noticed something about us. Our generation is definitely good at learning new tech, BUT we're also practiced at dropping something we mastered when the new thing shows up.

Take texting. I went through 3 Nokias between 1998 and 2000, all multi-tap. You tapped the 2 key until the A, B or C showed up, waited for the cursor to move, then started the next letter. It took coordination and a lot of patience. I got fast at it anyway. Then BlackBerry handed me a real keyboard and all that practice was useless, except for how quick my thumbs had become.

The Apple products I owned compounded my skills even more. iPod in 2001, iPhone in 2007, iPad in 2010, each one built on the touch habits from the last. Some of those skills died with the device. Others, like the touch habits, keep serving me well.

AI tools are like this just on a faster timeline. The AI you use today may not be the one you use next year and we all need to accept that. But the important thing, imo, is that the human skills you sharpen, like directing what you want with clarity and checking what comes back with judgment, stay with you.

The man who built that timeline, Ben Tossell, put the whole thing together in one morning with 7 AI agents and he is not a coder. A few years ago that was a team and a month of work. The devices keep changing. How you work with them, clearly and with a skeptical eye, is the thing that carries from one to the next.

PS You can build your own list here: bentossell.com/devices

Brett

Try This Today

Stop trimming every overlong answer by hand and set your preference once. In ChatGPT, click your profile icon, then Settings, then Personalization, then Custom instructions.

In the box for how you want it to respond, write something like: "Give me a short answer first. Go longer only when I ask for the detail." Save it. From then on it applies to every new chat instead of resetting each time. Claude and Gemini have the same setting under their own preferences.

Watch: why AI gives you an essay when you asked for a sentence

@aiover50

#creatorsearchinsights #aitipsandtricks Allie K Miller says Opus 5.5 is still too wordy to skim. She's right, and "fix it in settings" is... See more

Worth Knowing

A human has to review the decision when AI fires someone

California just made it illegal for an employer to fire or discipline you based on an automated system alone. A person now has to review that call and back it up with real evidence, like a manager's input or your own personnel file. You also have to be told when AI played a part. It was one of more than a dozen laws protecting workers from AI signed at once and the rules take effect in July 2027. Read more

The most wanted AI skill on Wall Street is not coding

Banks like JPMorgan, Citi and Capital One are hiring hard for people who can direct and supervise AI agents rather than build them. Demand for that one ability has jumped more than 1,700% and it is being called the hottest skill on Wall Street right now. The premium is on knowing what good output looks like and steering the tool toward it, which is exactly the judgment you have spent a career building. Read more

Is there really broad demand for AI Agents who can run your life?

The industry keeps pushing "agents" that would run your calendar, your inbox and your errands with little say from you and the pushback is getting loud. The argument making the rounds is plain: most people do not want a digital life manager, they want help with a specific job and then they want to stay in charge. Adoption of these agents that run everything is still a rounding error next to ordinary chatbot use. If handing your whole life to an agent doesn't feel quite right, you may be right keeping it as a focused tool. Read more

The creative part AI can’t fake

Pope Leo called on artists and the wider culture to push back on AI generated work, warning that the technology lacks the "spark of humanity" and that flooding the world with "powered by AI" output turns us into passive consumers of thoughts nobody actually had. I agree -- the human spark is the thing a model cannot manufacture. Read more

Tell me

Which device did you get really good at before you let it go? Hit reply and tell me. One line is fine and I read every one.