John Wilke’s Tooth //

//

[John Wilke’s Tooth // @JWTooth]

I bought my first @ParallelTCG NFT card over 3 years ago and have been in the community since. Through all of the amazing things the Parallel team has done, I have been most eagerly awaiting the physical cards. Well that day finally came and I’d say the wait was worth it…

[John Wilke’s Tooth // @JWTooth]

@templecrash @AIWayfinder @a_haldenby Goats

[John Wilke’s Tooth // @JWTooth]

Day 56 of opening a pack of @ParallelTCG a day to find a First Son of Mars.

I love the look of the Life Siphon 🔥

[Wayfinder Foundation 🧭 @AIWayfinder]

Rawdogging predictions = luck
Wayfinding predictions = edge

World Cup semifinals start today

Our Sports Agent finds you edges on Polymarket/Hyperliquid outcome markets vs. traditional sportsbooks

Your agent is a better researcher than you and is connected to the data that matters

[John Wilke’s Tooth // @JWTooth]

How did it take a year for X to figure out no one was seeing posts they actually wanted to see?

[John Wilke’s Tooth // @JWTooth]

Trying to speed these up a bit, attention spans are short

[John Wilke’s Tooth // @JWTooth]

Day 55 of opening a pack of @ParallelTCG a day to find a First Son of Mars.

Nice Uncommon SE 🔥

[Wayfinder Foundation 🧭 @AIWayfinder]

Introducing AI Loop Strategies in Wayfinder

Foundation model companies don't improve their products with one-off changes. They run loops that ship, measure, learn, refine, repeat.

We've taken that logic and applied it to trading strategies across tokens, perps, stock tokens, and prediction markets.

Wayfinder loops improve performance through:

1⃣Performance data - backtest, paper trade, live trade

Backtests show how a strategy performed historically and let you iterate with the agent.

Paper trading assesses it in real time with no downside.

Live trading is capital in market.

The agent uses the backtest as a benchmark, compares it against your paper or live results, and seeks to incrementally improve performance through these methods:

2⃣Exploitation - going deeper on what works

The agent continuously reviews the strategy's own data: what performed, and under what conditions.

If it finds the strategy thrives in high volatility and bleeds in low volatility, it indexes toward high-vol conditions.

This isn't a one-time optimization - the loop is always checking, always refining.

3⃣Exploration - testing what's new

The agent forms hypotheses beyond current trading conditions: is there an asset better suited to this strategy? An indicator that fits it better?

It backtests the idea and presents its findings. You decide whether to merge it. Human in the loop, agent doing the legwork.

A live internal example:

We gave the agent a theme: short IMX on downward momentum. We confirmed the opening parameters together, and the loop took over.

Every loop since has combed the strategy's trade data for the conditions where it performed, formed a hypothesis about an improvement (indicator, asset, parameter etc), backtested against the current strategy, and raised a version for approval only if it unambiguously wins. Currently earning 18% APR, and still iterating.

The strategy you deploy on day 1 shouldn't be the strategy you're running on day 90.

Now it isn't - it evolves with the market.

Public release coming shortly.

Not financial advice. Past performance ≠ future results. Digital assets are volatile — trade at your own risk.

[MetaMask 🦊 @MetaMask]

MetaMask just turned 10 years old. 🦊

To celebrate, we made you a little something:

https://t.co/RCUpaBF9eU

[John Wilke’s Tooth // @JWTooth]

New @AIWayfinder path 🚨

It examines the different ways to gain exposure to an opportunity.

It eliminates weak approaches using objective evidence, leaving only the approaches that survive scrutiny so you can make a more informed decision.

https://t.co/AABY1Q1xex

[John Wilke’s Tooth // @JWTooth]

@opensea @X //

[John Wilke’s Tooth // @JWTooth]

Took a week off with the family.. back to the grind. New @AIWayfinder path loading 👀

[John Wilke’s Tooth // @JWTooth]

@ParallelTCGpod Daughter was singing My Friends Over You in the car this weekend. It was a moment

[Easy @NotSoEasyMoney]

i genuinely cannot believe how spot on @AIWayfinder was with this trade...

I am DISGUSTED, that I completely overlooked it.

[John Wilke’s Tooth // @JWTooth]

@CoinMarketCap Don’t need to wait. Use @AIWayfinder

[John Wilke’s Tooth // @JWTooth]

@Berna7224 //

[John Wilke’s Tooth // @JWTooth]

@jeetsbyNav @AIWayfinder — I am also in my 30s with no time to trench and this makes it SO much easier

[John Wilke’s Tooth // @JWTooth]

Day 54 of opening a pack of @ParallelTCG a day to find a First Son of Mars.

My daughters favorite card, Change of Heart 🦋

[//Kalos @templecrash]

Not size but damn...

Those 400% + returns on @AIWayfinder using trench/meme coin tool for @RobinhoodCrypto are good...

[John Wilke’s Tooth // @JWTooth]

@CalerbCalamari @ParallelTCG @GoogleAI @Parallel_League Nice!

[CalerbCalamari @CalerbCalamari]

Updated my workflow for the @ParallelTCG Liquipedia page. Using @GoogleAI Spark agent to navigate the @Parallel_League Versus site and build out documents with the data on each event. Then I use a script to download the documents from my drive, feed the docs to Google Antigravity, which then formats them based of documents and examples.

Still don't have it perfect yet, but once this is running It will almost be a fully automated system for creating and adding pages to @LiquipediaNet. I imagine there is some way to link to the API and build the page itself too....but I figured the actual page building was a good place for a stop and human checks.