#334: The Bobu Magic Show and Trusted Execution Environments

Earlier today Azuki introduced the Bobu Magic Show, a livestream YouTube event on October 10th. If Bobu sounds familiar, I recently covered him a few months ago with his AI-enhanced YouTube livestream, and it looks like he’ll be back and better than ever.

This time, Bobu will be adding more tricks up his sleeve thanks to a partnership with Flashbots and TEEs (Trusted Execution Environment).

Flashbots is an R&D org focusing on minimizing the negative impacts of MEV (maximum extractable value) in the crypto ecosystem. The simplest way to describe MEV is providing a tradfi example: frontrunning.

Let’s say you’re a Robinhood user and trade stocks on the app. Bigger and smarter market participants will purchase the stock before you for a cheaper price and sell it right after. Frontrunning is just one type of MEV, and this happens onchain as well.

The Flashbots team does R&D work in other areas as well, and one of them is with TEEs.

A Trusted Execution Environment is a secure area within a device that isolates sensitive data from the rest of the system. A couple of examples that may use in our daily lives:

• Apple’s Secure Enclave: If you use Face ID, this sensitive data is stored here securely

• Android’s Trusty: Google Pay uses Trusty, a TEE

• Using the toilet: A bathroom is a ‘trusted environment’ that allows you to ‘execute’ securely 💩

The last example is one I pulled out of my ass, but is a decent ELI5. Conversely, this meme many of us have seen before is the opposite of a TEE lol.

Now that we know what Flashbots does and have a high-level understanding of TEEs, what does this have to do with Bobu and next week’s magic show?

Earlier today Bobu’s account posted this:

Nothing crazy, a regular old post right? Not quite, this was drafted and posted by DGTL, an Azuki community member who didn’t have direct access to the Bobu account.

However, he did receive a NFT that granted permission to post once on the Bobu account. The NFT incorporated a LLM agent that set parameters around what content could be posted (eg: no swear words, etc.) and how many posts could be made. Once the NFT was used up it was burned. Azuki provided the transaction showing the NFT burned once posted.

It isn’t a magic show without volunteers, right? Azuki’s announcement encouraged the community to participate. In order to sign up, participants have to authorize the Bobu app on X and create an account link allowing Bobu to post on the volunteer’s X account.

And based on DGTL’s example and AI livestream Bobu, there may be some entertaining posts that come from the show.

If you want to try out this tech, check it out on Teleport.

The implications of this type of TEEs allow digital experiences to become much more interactive and involved. A simple way to think about this is with beer pong (lol bear with me for a second).

TLDR beer pong is drinking game where you set up red cups, shoot ping pong balls into the cups, and drink beer whenever your opponent makes it into the cups. I played many beer pong games during my college days and lost most of them because I sucked.

One fun feature of the game is the ‘celeb shot’. Friends at the party could take a celeb shot on your behalf and hopefully make it.

What DGTL did on Bobu’s account was the digital version of a celeb shot, but much more dynamic. How could this be used?

• A project account could provide access for a community member to post once (or more) on their behalf. This is the DGTL posting on Bobu account example

• A community member could provide access for a project to post once (or more) on their behalf. This is what volunteers are doing for Bobu’s Magic Show

• [Hypothetical] Fantasy Top Heroes could allow one of their holders (via token gate) to post on their behalf as a perk

• [Hypothetical] A gaming streamer could let an audience member or friend play on their behalf for a game while continuing to livestream

• [Hypothetical] An influencer can conduct a live giveaway, but delegate the execution of the giveaway to an audience member

What makes this interesting is that these real and hypothetical examples don’t involve credential (password) sharing and have specific parameters and safeguards set.

Kudos to the Azuki team for their continuous experimentation around interactive experiences. I think they’re onto something here, and it may become a new way for experiences to become even more interactive.

Do you know what a TEE is now? If so, share or subscribe!

Shape launched their creator-focused Layer 2 blockchain on earlier this week. Accompanying the network launch was the introduction of Stack, Shapes’ version of online reputation.

How does the Stack work?

• A Stack contain onchain achievements and users earn medals by interacting with and contributing to projects on Shape

• There are 4 medal tiers: Special, Gold, Silver, Bronze

• Each medal doesn’t have a specific value. Instead, leaderboard ranking is based on the number of medals a user has earned, with the higher tier of medals being more valuable than any number of medals from a lower tier.

• Medals can be revoked if a user no longer fulfills the conditions required to earn them

• You can set up your Stack here

The easiest comp for this, and potentially where Shape drew some inspiration from is the Olympic medal count. There are two ways to rank medal counts, by total medal count or by Gold medals.

In this comparison Japan is 6th in total medals. However, when ranked by gold medal count, Japan is bumped up to 3rd since any number of gold medals is ‘worth’ more than any amount of silver or bronze medals.

Shape’s Stack leaderboard follows a similar logic, with Special medals being the most valuable.

Although the Stack leaderboard is only a few days old, we can see how this type of leaderboard system (as opposed to a points-based system) ultimately leads to the same type of desired user behavior while having a more refreshing approach versus the standard points system.

When it comes to perks and benefits of being at the top of the leaderboard, Shape also takes a different approach. The first incentive was that top 250 ranked users had a chance to win a mint spot for DeeKay’s DeePle project. The higher a user’s rank, the higher the chance they would win a mint spot.

How did it work?

• The top 250 leaderboard snapshot was taken on Oct 3, 4:04pm UTC

• Random numbers (via Verifiable Random Function, VRF) were generated for each address in the top 250

• The output logic to determine who won the allowlist was Leaderboard rank x VRF number. The lowest final outputs win

  • If Rank 1, Leaderboard multiplier is 1

  • If Rank 100, Leaderboard multiplier is 100

  • If Rank 250, Leaderboard multiplier is 250

• Shape details their methodology here along with the onchain results

So if you were ranked high in the leaderboard, you were highlyyy likely to receive an allowlist but not guaranteed. Makes sense considering the fact that the top-ranked users should have a significantly higher likelihood of winning the perk compared to the lower-ranked folks.

It’ll be interesting to see how Shape continues to take a unique approach to reward top collectors and determine who wins. At a minimum, it’s nice to see a different version of a leaderboard appear in a sea of points systems.

See you next week!