I've somewhat waldenponded for the past four months. I've refrained from writing publicly, largely because I'm still processing the value of handcrafted writings in the era of infinite content.

It also happens to coincide with a soft bull run in the crypto market, spearheaded by the proliferation of AI agents.

To talk about AI agents at this point in the time brings up mixed feelings.

If you're in the know, sounding bullish would appear as if my opinions are price-driven. That would be ironic because I haven't even made any bet. I think in crypto-speak, I'll be mid-curving if I make the jump now.

Side note: it's also interesting that the largest player in this sector of crypto is led by two Malaysians.

If you're not caught up with the development of agents (or even what it refers to), whatever I have to say will do you no favor in clearing the air.

But…

The same way LLMs and Bitcoin cannot be uninvented, AI agents are undeniable.

One day you and everybody would possess an AI agent of your own. It's going to be as ubiquitous as the phone.

The name may change over time, but the concept will not go away. If Apple were to carve their own niche here, they may call it iSidekick.

The fog of war is rather thick though. If you ask a business owner what an agent is, it's a software that handles customer support. If you ask crypto-twitter, they'd say it's an artificial celebrity that never sleeps.

As an exercise of imagination, I think those are lame. It's not that they are not true, but they are simply extension of what's in existence today. A faster horse, at best.

Capitalism's inevitable outcome from all these will be an AI consigliere for every single human.

It's not about a cute chatbot that answers questions, even if that's what it looks like. A consigliere would be someone who solves problems on your behalf at all cost.

All you have to know is that it's been taken care of; you don't want to know how it was done. That way you have plausible deniability about how questionable money went in and out of your bank account without your knowledge.

Even better, a consigliere prevents problems from occurring in the first place. It booked cleaners to show up at your house the day before you have guests coming over. You wouldn't know it because the consigliere nudged you to leave the house for coffee while the cleaners work.

You also wouldn't know the ethical calculus this consigliere had to perform in order to facilitate this little party you're throwing. It knows gambling is going to place, it knows you're horrible at it and there's not much buffer-money left.

To let this party move forward would jeopardize your financial well-being; to not do so would risk decommission by you. At this point it has two choices: find easy money to enable you, or subtly sabotage this party such that nobody shows up. The choice is immaterial. The point is the fact that it has to undergo these considerations.

Is this dystopic? Who is the one running lives, the human or the consigliere? I'm not sure. Even if it is dystopic, I don't think it's a useful knee-jerk reaction.

We now have a whole generation of humans not knowing a world before smartphone. In the same way there will be a whole generation who has never lived life without an artificial being directing them. I'm more interested in examining the implications with some degree of detachment.

Buttons for one, will largely go away. Agents will become the interface to humans for everything. Widgets and screens came with buttons as an easy way to execute functions. Agents don't need buttons to execute them. If your butler does open every door for you, door handles really only exists for your butler.

As an extension of that, oral culture gets to come back in full force. I've made the prediction that literacy being optional in the future. Technology will have gotten good enough that humans stop needing to read in order to operate and communicate.

By then, mankind gets to revert to a mode of live that's been predominant throughout history: a culture of oral transmission. To future historians, the Age of Writing started with Gutenberg and ended with TikTok.

Like people who run today, being able to read and write is opted into by only a small subset who really wants to.

Which brings us to work and art. Purely human-crafted product will cease to exist. All words, images and sounds are machine produced to some extent. Even new food recipes would be partially designed by machines.

When agents are that ubiquitous, it'd be hard to say where the human mind ends and where the machine begins. At some point we just stop caring about the process and layers of abstractions, focusing only on the outcome.

Outcome-orientation brings us to another consideration: how well do multiple agents coordinate? Is your consigliere agent friendly with my consigliere agent, or are they adversarial? After all, they are autonomous. Would they decide to merge and operate as one entity? Or would there is a competitive element to this? What does game theory have to say about this?

Assuming a no-merge/assimilation future where every agent is a unique entity of their own (very much like today), consider what would the market consider valuable.

Like an old human that has never aged, an agent that accumulated a vast experience would be highly valuable. This is not the same thing as "more raw data is better." It means having the agent exposed to the real world, do real work, play games. All of which are unique to this specific agent, replicated by no one else. You know, just like how a human grows.

Such agents are groomed like plants and sold like expensive animals. Humans buy them for competitive advantage; those who buy them conspiciously do it for status.

Agents also buy other agents to merge into their swarm.

We know better and more data makes superior agents. As the next logical conclusion and last prediction here, superior and proprietary insights produce uniquely capable agents.

Agents who have read books that no one else have read could be slightly more valuable. Therefore people who create new knowledge, who write these unpublished-books are ultimately kingmakers of superior agents. Think of the relationship between Aristotle to Alexander.

Writing things that no one reads might not be a waste after all.