0:00
/
0:00
Transcript

What Is The Point of AI?

pattern matching you to opportunities you’d realistically act on

But pattern matching: that’s the easy part.

Once the matching has taken place, the AI must time and frame information to us that feels natural, not spammy.

Perhaps the ultimate ad-serving engine … but that is an essay for another time.


We are in early days of this frontier tech so the value in just pattern matching is overwhelming enough on a daily basis: omg I can use AI for this, that, and this and that …

But we are already publicly fatiguing from the inept use of AI by other humans: AI Slop.

Yet, it’s not really the AI, is it?

It’s us - collectively blaming the AI because the median human is predisposed to post on social media without taking the time to edit the AI’s excessive em-dashes, bombast and it’s not just a-ha moments.

Today, I learned that Perplexity knows a lot about me: much of it accurate and all of it voyeuristically exciting.


Conversation with Perplexity

Here are some snippets from our conversation:

Perplexity, how can I get USPS to support my Li-ion battery recycling scheme?

I wasn’t asking about coffee so I said …

Perplexity, are you clocking me?

What else does Perplexity know about me?

I told Perplexity to fuck off like a Brit would

Unfazed, I dropped a bomb on Perplexity …


Perplexity, behave! PS - I did not work at Facebook, don’t teach yoga and much prefer JLCs.

I pushed for more information …

Sensing opportunity, I pounced …

We exchanged some more details about this and that before Perplexity composed a Weekly Task for itself.

I can’t wait for my first weekly update which drops tomorrow.

Reflection

Our conversation drew me in and despite my initial frustration of being watched I quickly moved to what could be in it for me. How might I use this tool to my advantage: not to lord power over others, simply to be a better version of myself.

I learned about Perplexity Tasks and how to set them up.

Is there a risk that my deep learning imprint is being shared with my AI’s overseers? Probably yes and on a meta-level it’s guaranteed to be. Our collective usage of AI is enabling optionality for the AI to get better, even if it is not a linear process.

But I think this point in time is different.

Specifically, we have been trained to distrust the Internet Man over the last 25 years as we have become more aware of social surveillance and its winners and losers.

The power dynamics of Internet ad-tech were always asymmetric, even though there was a mutual value exchange: free product and you’ll be the product.

At this specific point in time (2026) the power dynamics of AI are much less asymmetric and therefore you can and should use this to your advantage.

Leader or Follower

You can live a great life either way. But you will cause yourself much more stress in the process if you don’t admit to yourself which one you are.

You can adapt your life at any time to move from one category to the other. Life requires you to do so sometimes.


Music: Night Swim by bahkin

#companyman

Discussion about this video

User's avatar

Ready for more?