johnmaconline

Could AI-you offload a bunch of administrative tasks? Could it manage your schedule (like you would) and the supply chain required for running your household?

“Hey, AI-me, as you know, we’re having a party next weekend. Here’s the menu. Please ensure we have the ingredients on friday.”

“Hey John, your Mazda is due for inspection at the end of the month. I’ve scheduled with the garage for Wednesday of next week because it was the only day they had available. I needed to move the lunch plan you had with Jack to Thursday. He was good with that.”

Could AI-you deal with household issues quick and easy?

“Hey, AI-me, the garage-door just broke. Get someone here ASAP.”

“Hey John, the hot-water heater started leaking during the night. The plumber will be here in 35 minutes.”

Could this AI-you manage your finances? The day-to-day and also investments, find money-making opportunities (specific to you, your skills, and view of the world), capital improvements, etc.

“Hey, AI-me, we’d like to like to redo the office this winter, but we’re wondering if a new fireplace and wood flooring would fit into the budget. Could you work up a bunch of options based on these ideas…”

“Hey John, you should look at this opportunity I just found. It looks like it would fit your skillset and might bring in an extra $1000 a month.”

If AI-you could do those kinds of things, would you want it to?

Elon thinks Optimus is the perfect household assistant. He envisions a future where humanoid robots are performing many of the tasks that we spend our day doing today — cleaning your house, cutting the grass, tending the garden, grocery shopping, and walking the dogs.

I disagree. I wouldn’t have it do any of that. But what I would do is walk it across the United States. Drop it off on the Ocean City Fishing Pier, pat it on the butt, and say, “I’ll see you at Pescadero Point.”

What a cool engineering and social experiment.

Think of all the technical hurdles it would have to autonomously overcome to make that trip — recharging, repairs, weather, terrain, navigation, etc.

Think of the social curiosities it would generate and hurdles it would have to navigate.

If you saw an Optimus just walking down the street, what would you do?

Would you engage with it? Would you ask it questions? What if it looked ragged or broken? Would you help it? Would you knock it over? Would you invite it to your home? Would you play games with it? Would you encourage it? 

I wonder what Optimus would learn by walking from Jersey to California.

More creatives are signing open letters to stop training LLM models on their content without proper licensing (i.e., financial compensation).

Fundamentally, this is a legal copyright issue (i.e., money). Not a “stop the AI from taking over!” issue. 

Licensing issues (i.e., bucks) aside, would you want an AI version of yourself? (Note: A ChatBot version of you).

Well, what could it do for you?

Picture a 10-minute morning meeting with your AI-you each day. Of course, it’s been trained on all of the previous days of you, so it knows your history, how to sound like you, your preferences, and how you generally respond to the requests, challenges, and tasks of the day. But the purpose of the morning meeting is to understand the today-you. It asks you about how you’re feeling, things on your mind, what happened yesterday, and stuff that is important (as far as you know). 

Presumably, this AI-you could stand-in and take your place in particular aspects of your daily life that don’t require you physically. Let’s also assume that AI-you really is a good version of you. 

Maybe, in some ways, a better version of you because it’s you at your best, most consistent self. Rarely, if ever, would you review what AI-you did and think to yourself, “that’s not how I would do it” or “I’m not happy with that.”

But I don’t think this comes down to pragmatism.

I think it’s simply, “Would you want an AI Version of you?”

If the robot can drive the car, should you?
If the robot can harvest the corn, should you?
If the robot can serve the drinks, should you?

But what if the robot can negotiate border disputes? Should you?
Or what if the robot can decide value? Should you?
What if the robot can prioritize? Should you?

In the near and far future, robots can do many things.

With humans, it’s not about can. It’s about should.

It’s easy to type the code.

It’s hard to figure out what the code should do. That’s also true for writing books, creating value propositions, and developing spreadsheets.

The typing is usually the easy part. AI is a very good typer. But the value is in the imagining, figuring, and designing. That’s where the creativity and choices reside. That’s where the risk resides.

AI doesn’t take on nor understand risk. Oh, it understands risk as a math equation to be evaluated. If you can design an equation for risk, AI can evaluate that risk all day, every day.

But it doesn’t understand risk. It has no emotional connection to risk.

You do. Make no mistake. You’re the one taking on the risk.

That’s why you have the value.