What If AI Brought Opportunities to You? [Raw Session]

00:00 — Waiting for someone to notice

Hey, welcome back to Slow Builds. So lately I've been thinking a lot about how much opportunity still depends on somebody noticing you, and whether AI could eventually change some of that.

You might send somebody your work, apply for a job, or upload a video and hope the right person sees it. Sometimes a connection comes through somebody you know. Other times somebody just happens to be there when you do something very well.

We've built all these systems for finding people: recruiters, scouts, auditions, applications. And they all work to some degree. But being capable of something and getting a chance to do it are still two different things.

You might never get in front of a person who would recognize what you can do. You might not even know there's something worth showing them. That's the part I keep tracking back to and coming to.

And what if AI could help make those connections from things you're already doing? Something you've chosen to use that can notice a possibility and bring it to you. You still have to decide what to do with it, but maybe you won't have to know exactly what you were looking for before anything could happen.

01:20 — The kid running for the bus

Now, the picture that started this in my mind, and what made me think about it, was imagine a kid running because he's late for the bus. He's not trying out for a track team. He's just trying not to miss the bus and be late for school. But maybe he's quicker than he realizes.

And unless somebody happened to notice, he gets on the bus and his day continues as normal. Or maybe you miss the bus and you run the whole way, which is another thing. Maybe he's a marathon runner and not a sprinter. But like I said, he gets on the bus and he doesn't think anything of it. He just had to catch it. There's no reason to think much about it.

Now, one run wouldn't tell you whether somebody has unusual ability. And I'm not imagining cameras everywhere, judging kids on their way to school. But suppose he already uses something that tracks his activity, and he and his family chose to let it look for patterns in his running over time.

Maybe it notices something worth exploring. It wouldn't need to announce that he's going to the Olympics. It could just ask whether he enjoys running and help him find somewhere to try it. Maybe there's a local club he didn't know about. That's a small connection, but without it, running might stay something he only does when he's late.

And I think that same gap exists in plenty of less obvious places. Because as unobvious as that one is, there's even more that are less obvious, I believe. And that's just a simple example of someone wearing a tracker. They've got their Apple Watch, their Fitbit, or their WHOOP or something that their parents put on them to track them or to notify them or keep access or whatever. And it just happened to have an ability attached to it that could look like, “Hey, this is worth looking into.”

03:17 — Abilities we might overlook

And so in my case, somebody could be making videos in a tiny little room with almost nobody watching them. Sounds very familiar. Maybe there's something in the editing or the way they explain an idea that could be useful in work, something they never considered. They're thinking about making their next video. They might have an idea that somebody elsewhere is looking for help doing that exact same kind of thing.

Or think about a person at work who's always figuring things out when something breaks. They might—back to the video, I like my ideas. I just don't think I'm the right person to be putting them out there. Maybe I am, but I don't think I'm the right person to be editing in any way whatsoever. I do a one-take shot and I'm done.

I would like to learn and experiment more, but I don't know. So maybe there's someone out there that might eventually find my videos, or there's a tool that lets people find other people making videos, and they can help them, I don't know, team up. I'm just thinking outside the box and way off script when I do this.

But again, bringing it back to that, somebody could be at work right now, and they're the person who always figures things out when something breaks. They might not have the right degree, the right experience, or know anyone who could help them move into another role, or move up in the role, or switch jobs completely. They might not even describe themselves as a good problem solver. It's just what they do.

If they chose to bring examples of what worked to an AI, could it help them see where else those abilities might be useful?

And that interests me because we don't all get the same chances to be noticed. Where you grow up, the school you went to, and who you know can affect which possibilities you even hear about. Some people also have more time and freedom to put their work out there. Other people are busy getting through the week. AI wouldn't remove those differences, but maybe it could help someone find a connection that currently depends on knowing the right person.

05:28 — A first step that fits your life

The question then is what happens after it finds one. Because telling that person at work they're a good problem solver doesn't give them much to go on. They might say, “Okay, thanks,” and go right back to doing the same thing. What's the next problem on the list? Give me the next ticket.

What kind of work could they explore? What would they need to learn? Would they even enjoy doing it all day? That's where I'd want the help to continue. Maybe there's a small project they could try, or someone they could talk to who actually does that work. Something that lets them find out a little more before committing to a big change.

And that first step has to fit their life. If they have a mortgage and a family depending on their paycheck, a suggestion to leave work and retrain for two years might not be something they could use.

The same goes for a kid and a running club. Finding a club is useful, but maybe the fees are too high, or no one is available to drive them there. So the next question would be whether there's something nearby or support the family [unclear] know to look for.

For the person thinking about different work, it might mean finding a way to try it without leaving their current job. Those details are part of the opportunity. Without them, we could get much better at showing people possibilities they still can't act on.

I don't necessarily need an AI to tell me I have some hidden talent. I want it to help me find something worth exploring and a first step I could actually take. Don't just tell me I'm a great runner. Tell me why I'm a great runner. Give me opportunities and suggestions on how to train, where to train, and who to contact.

Same with the work. What can I do at my current job, or do on the side at home, or in a short period of time? What tests could I take? What skills could I take? Give me a problem to solve.

07:26 — Useful help needs personal context

But helping with those details would mean knowing something about the person. It might need to know roughly where they live, what time they have available, or what they can afford. And this is where the idea starts getting uncomfortable for me, because now it's getting into privacy.

The more useful I imagine it becoming, the more information I can imagine giving it. Something tracking that kid's running could also end up keeping a record of where he goes. And an AI helping someone understand their work could be given conversations or details about things they're struggling with.

We could start with a fairly simple request and gradually build a very detailed record of someone's life, all because each extra piece of information seems like it might make the help a little better.

08:16 — What I share should stay my choice

Choosing to use it would have to be the starting point. But agreeing to that first request couldn't mean agreeing to everything that follows. I'd want to know what it actually needs for the thing I've asked it to do, and whether I can still get useful help without giving it access to everything.

For me, that would mean being able to decide what the AI gets to see and what it's allowed to remember. If I ask it to look at a few pieces of work, that should be the scope of what I'd agreed to. If it needs something else, I'd want it to explain why and let me decide whether to add it or not.

I also want to be able to stop and delete what it had kept, and choose whether anything gets shared. Finding a possible connection for me wouldn't give it permission to send someone else a profile of me. I might want to explore an idea privately before telling anyone about it. I might decide to leave it alone altogether and not even pursue it.

And I wouldn't want information I shared for help quietly turning into a report for my employer, an advertiser, an insurance company. If it's supposed to be working for me, I need to decide what happens with the things it notices.

09:29 — Being good at it doesn't mean I want it

And that includes being able to say that it's noticed something real but drawn the wrong conclusion. Maybe I'm good at organizing people because I've spent years having to do it. An AI might see that and keep finding me opportunities to manage people. But that could be the part of my day I'd most like to get away from.

Yeah, I don't like meetings. But I'm okay with small groups. I'm okay with a great team. But being responsible and reporting back up is not always the funnest thing to do.

But like I said, it might know I'm great at something, but it doesn't know that that's the thing I don't like doing. Because being good at something doesn't mean I want more of it in my life. And what I've done so far isn't necessarily what I want to do next.

If I spent ten years doing one kind of work, there's going to be plenty of evidence that I know how to do it, and probably know how to do it very well. There might be very little evidence for the thing I'm interested in learning. I might be fairly bad at it right now. I'd still want room to explore.

So I need to be able to tell the system what I'm trying to move toward, even if that doesn't match the patterns found in my past. And I want that answer to actually change what it brings to me. Otherwise, I could end up with something that knows a lot about what I've done but keeps missing what I'm asking for.

Maybe there's something I dabble with on the side, and because I only get to spend very little time at it, I might be getting worse and worse, and my ability looks really bad. Or maybe there's something I do spend a lot of time, too much time, my extra time, and I really have a passion for it. But maybe I'm really, really bad at it.

So there's all these ins and outs, and I need honesty, and I need truth, and I need to give permission to certain things, but I need to control it all at the same time.

11:28 — When AI gets it wrong

So, of course, it could also be wrong about what I'm good at in the first place. It might mistake a few examples for a pattern, or overlook something because it doesn't recognize the way I do it. And I'm worried about whether certain people keep getting shown certain kinds of work. You could think you're opening up possibilities while quietly narrowing them.

A suggestion would have to stay something I can question and test. I wouldn't want to treat it as a verdict on what I'm capable of, especially if other people start trusting those judgments.

Because once getting an opportunity depends on what an AI thinks of you, there's a reason to start performing for it. You might choose work because it looks good to the system, or change how you present yourself because you think that's what it rewards. Then we've added another audience you have to impress.

And that brings me back to why I like the idea in the first place. I was imagining someone getting on with their life, with a bit of help finding possibilities along the way. I wouldn't want ordinary life to become one long audition.

12:32 — Who is the AI working for?

So I keep coming back to who the AI is actually working for. So much of my experience with algorithms has been about what they want to show me next: another video, a post, something to buy.

What interested me here is being able to ask for help finding things I actually want to explore. That would involve listening to what I want, accepting what I don't want to share, and understanding that I can turn something down, even something I might be good at.

I don't know how well that would work. There are probably problems with it I haven't even thought about. But I keep thinking about that person who's capable of more than their current situation has given them a chance to try. Maybe they don't know what to apply for, or who to talk to, or that something they already do could lead somewhere else.

I can see how a useful connection could make a difference there. Something they've chosen to use notices a possibility in what they're making, learning, or figuring out in their ordinary life. And it asks, “There might be something here. Do you want to see where it goes?”

I like that possibility. I'm not sure how we build it without slowly turning it into something else. I'd want the person getting that suggestion to still feel like the answer belongs to them.

13:53 — Closing thoughts on discovery

That's why I like the idea of silent tracking, silent noticing, by maintaining privacy and protecting the person's information. But finding trends, finding oddities that the person is showcasing or doing. And you can make suggestions. You can let them know, “Okay, you're great at playing this video game. Let's try this one. Oh, there's a tournament for that. Would you like to join for it?”

Or the running thing. I love that idea because it's so simple, and it's so easy for a person who doesn't run or do any kind of exercise, like swim or row or anything, bike. And once they have a tracker or they pass by something, it can pick up on that. And it could give suggestions. It can ask if they want to pursue that.

And I think there's ways to add nudges to this, add metrics that are private. But also, for the person that it's giving the suggestion to, it gives them an opportunity.

I think the world could become a little more open, a little more of a place that we can all explore more. Because we all get bored on a rainy day, or on a day when we don't know what to do. And we don't like wasting our time. We don't like just watching TV and binging on social media.

So going out, or going online to talk to other people, or explore new things, or try new things. I think AI has a great way to present opportunities to see where we can grow and become great at what we don't even know we're great at.

All right, thanks for watching.