Wai Oh Wai?

Not a day goes by without Cool New Stuff™ being churned out to satisfy the needs of degenerates like you and me, something very reminiscent of the early days of DeFi Summer.

There’s a big difference this time around though: most projects have no VCs.

After years of large VC rounds prior to token generation events, retail has been left with scraps and a price chart stuck in FDV Hell going down and to the right.

In the current era, retail (we degenerates) can get in on the ground floor. Here, we pause to pay our respects to $YFI, the original DeFi immaculate conception.

…

There’s an opportunity: to approach new AI Agent launches in a smarter, systematic, and ideally, alpha-capturing way.

The same niche was filled in the off chain world by Y Combinator. Seed/angel investing had been an informal set of word-of-mouth networks that traditional VCs largely ignored. Y Combinator filled the void in an intentional, methodical way to the result of hundreds of billions of dollars and changed the startup investing landscape forever.

Is there an analogue in the Agent Awakening? WAI Combinator is going after that same niche in the AI Agent space today.

It’s a mighty challenge. As the person running the Twitter account says:

Success will require the ability to deploy capital wisely in a cloud of uncertainty.

So far, the results aren’t terrible.

Claude Asks Me About My $WAI Investment

Do you have a position? Please disclose your bias.

Yeah, I got a bag at the mid-30Ms market cap for my personal account.

Do you have other similar positions?

No. I have large concentrated positions in $VIRTUAL and $AI16Z. By grabbing a bag of $WAI, I’m taking a Barbell approach: the $1B+ large cap leader platform plays on one end of the barbell, and the microcaps on the other.

Additionally, everyone is buying the infrastructure plays, which means there’s an opportunity to win on agents. I’m not smart enough, nor have enough time to evaluate each new agent project, so I’m outsourcing that to $WAI. I don’t like the “medium size” part of the market. My audacious goal is to make a 100x on this bag.

Is this a position for the gm Portfolio?

Not yet. gm Portfolio is a partnership, and the partnership is still evaluating.

Why not other projects in the same niche?

Dunno, I like the vibe of the person running the Twitter account and the clarity of his thinking. Bits of wisdom have come through like:

In the old days before medical imaging advanced to the point where it stands today, surgeons would have to make a decision on whether to take a patient for surgery for suspected appendicitis based purely on history and clinical exam. They targeted a 20% false positive rate, i.e., they knew they would get it wrong and take someone without appendicitis to the operating room 20% of the time. But anything less than that, and they were delaying treatment for people who had appendicitis resulting in worse clinical outcomes, possibly life threatening some of the time.

Indeed, the approach in the tweet above can be generalized to other parts of life: if you’re not failing enough, you’re probably not aiming high enough, and as a result, capping your upside. Lean into taking action in a sea of chaos, and get comfortable taking losses some of the time.

What do you know about the team?

Not much.

Are AI Agents actually involved in the investment decision making?

Not sure. I think I read somewhere it’s a mix of humans and agents. But I can’t find it anymore.

Could it be a scam or rug pull? Isn’t the lack of links on the project’s Virtuals page suspect?

Dunno. Could be I guess.

Could it go to zero?

Sure.

Aren’t you being reckless?

Possibly. But like I said above: investing in any project below $50M FDV involves deploying capital in a dense fog. A large part of this investment comes from my gut. My gut has been wrong before.

Aren’t most Agents going to zero?

Probably. But some will go to billions.

Is The Tinkering Society building agents?

Some members of the Society are, ahem, tinkering with agents, but nothing official. The Society is focused mainly on Size Credit at the moment, along with a top secret information market project. We love AI Agents on chain but our priority at the moment is with fixed-rate credit markets.