Shapeshift into the Agentic Future

ShapeShift DAO
1

Mission:

The world’s borderless financial system shall stand on open, uncensorable, decentralized protocols. We shall be the pathway into this borderless financial system.

One year Vision:

By 2026 The DAO evolves from Layer-1 swaps aggregation to supporting AI-native wealth creation - where your assets work for you through an intelligent agent swarm, without the friction of wallets, powered by collective intelligence. A stretch goal would be contributing to or shipping a real-time adjusting model that everyone can opt in and contribute to. We may not reach all of those pieces in tandem, but the evolution towards innovation is important.

Why This vision matters:

It’s been a long time since Shapeshift pushed the space forward by innovating. We need to get our name out there as a DAO of builders and get apps into users hands that they love/find useful. The memetic truth about chains is you optimize for creating hype, farming yield, or hosting stablecoin rails. Thus, we should gradually phase focus on something that is sustainable and diversifies us from one product.

If everyone is zigging towards the wallet powered perps casino we should zag & optimize for performance across chains, making their unification seamless, and aim towards friction-less day-to-day experience(s). Eventually creating wallet light, and walletless UX. The more we push the complexity under the hood for users – while maintaining their control– the easier it is for them to automate their holdings, or manage the fast moving hops through hype and/or yield. It’s up to them!

The iterative app differentiation strategy

The path here is straightforward enough, and hopefully feels less wandering:

  1. Improve the current app more judiciously.
  2. Continue to aggregate the best of Defi without over burdening the stack. Cut just as much as we add.
  3. Ship two/three apps of increasing size & utility and gradually pivot.

These apps are break down into three high level sprints, with flexible timing where we can mix/match features & use cases.

Now (Next 30 Days): ShapeShift OG

  • Bring back the OG experience and engage as many OG’s as possible
  • Goal: Immediate awareness bump and consistent walletless swaps.

Next (~Days 31-75): Chat with the sweeper

  • Launch POC agent with sweeping assets as a primary use case.
    • If we can do something visionary like passkey-protected consolidation, or vultisig TSS that’s great.
    • Currently debating bring-your-own-wallet to start and not build out wallet infra into the agent. There is a lot of risk there and weird rabbit holes of managing keys that we should avoid.
  • Multi-wallet sweep functionality. Chain by chain.
  • Leverage Eliza OS or the Rig
  • Use as a foundational AI proof of concept leveraging our current infra.
  • Goal: Demonstrate wallet-free future and generate large volume from new users.

Later (~Days 76-120): Agent Infra Foundation & iteration

  • Research, deliberate, and deploy more AI agent infrastructure.
    • Continue fine tuning research on private RAG and inference models for users, or best providers to cobble together (e.g. Venice, bring your own API keys. Bring your own server/DB/model/compute?)
    • Smart contract wallet or multisig infra extensions.
    • Presuming an agent is a wallet under the hood you chat with… complete research on handling human-in-the-loop custody, private key in users control, tx batching, and multichain intuitively.
  • Goal: Most decentralized & private mover in AI-native DeFi.

**Eventually ** (~Days 120 - unknown): Build a model that everyone owns, host it on chat.shapeshift as premier venue, for sane revenues or cost-sharing with users.

  • Expand distributed alpha protocol architecture research.
    • DisTro by Nous is promising as is Psyche by nous. Their research seems to be going in the direction we imagine.
    • Numerai was directionally correct but early. It’s use case should actually be training models, inference, and using cryptography obfuscation on the proofs.
  • Connect with in person teams at the bleeding edge and hack together on whitepapers and prototypes.
  • Goal: Reinforces above and allows Shapeshift to build a protocol. “Seti at home for alpha.”

1. what ???

To make it more clear, these next sprints/mini-apps solve the following problems:

  1. Wallet Friction Dies
  • Users hate managing multiple wallets
  • Passkeys and/or threshold-signature-schemas are the future of UX abstraction.
  • We can make interacting onchain seamless
  • Two apps that involve minimal asset selection and overwhelm is good
  • Pushing ahead into AI is not only well timed in the market, it’s also the future of all of our jobs.
  1. Assets Become Intelligent
  • A portfolio gets an AI brain.
  • Natural language instructions allow for fine tuning, Individual knowledge graphs for control.
  • Users “owns their edge” due to custom context windows, fine tuning, and inference you can manage.

  1. We get some exposure and try new things. Crypto :handshake: AI are very powerful and we get to be on the bleeding edge.

  2. Community Edge Compounds (if we build it)

  • Real-time alpha execution encrypts to a distributed model that updates in parallel.
  • Pure reinforcement learning, combined with a distributed zero knowledge system creates a model that everyone owns.
  • Share verified performance without giving up your balance, trades, or any specifics.

Risks & mitigating them

There are no bold moves without significant risks. A few categories come to mind, including Technical & Infrastructure Risks, Resource & Execution Risks, Trust & Adoption Risks.

Mitigating these continues our approach of smooth iteration, but it doesn’t mean we can’t raise the bar each time.

To mitigate some risks:

  1. Start with simpler AI integrations that add clear value (like natural language in the trading interface) or the smallest chunk of a “persona” that “recommends” actions to users.
  2. Leverage your IRL connections and community learning to build authentic expertise & partnerships.
  3. Focus before expanding to “agent swarms.”
  4. Build modular systems so components can be reused across mini apps.
  5. Consider strategic partnerships with existing AI infrastructure providers (Venice?)
  6. Continue evaluating a DEX partner that would allow an escape valve of advanced order trades on more than one chain/platform. (RUJI, Hyperliquid, TBD)

User research doc again

What Success Feels Like (qualitative)

  1. Community & Ecosystem Impact
  • Shapeshift becomes the go-to platform for AI-native DeFi activity
  • Resuscitating community dev involvement into an active developer community building on top of our agent infrastructure.
  • Recognition as craftsmen in the overlapping fields of wallet-less DeFi and AI.
  1. User Experience Transformation
  • Users view ShapeShift as their trusted gateway(s) to onchain finance. The place to “shift” onchain.
  • Portfolio management becomes as simple as having a conversation.
  • Community-driven intelligence becomes a genuine competitive advantage.
  • The narratives around the DAO and brand from 6 years ago, fade away.
  1. Cultural Achievements
  • Return to our roots as bold innovators while maintaining decentralization
  • Strong builder culture with rapid experimentation and learning
  • Balanced approach to innovation and stability
  • Stacked bar chart of revenue

The future of DeFi is intelligent, accessible, and collaborative. ShapeShift aims to be at the forefront of this evolution, building the infrastructure that makes financial freedom a reality for everyone.

What success looks like (quantitative)

Teams will refine and flesh out these rough goals into a more formalized post:

  1. Financial
  • 400K MRR (aka 200MM volume & 400K monthly recurring revenue)
    • :point_up: this is just a 10x from where we are now!
    • it’s unrealistic as it’s 84% growth YOY, but lets aim for the stars. We’ve seen 10% MOM growth from Dec 2023 so no reason to assume we can’t get there.
  • 30% conversion rate from land > swap
    • We’re currently at around 3%. Much work has gone into measuring this and we’re also operating of mixpanel which captures about 50% of the actual data.
  • $50K revenue from smini apps (smol+ mini= smini) monthly.
    • Once we capture figures monthly we aim towards more growth.
    • Ideal use case is 1000 users who love the proof-of-concept. We’ll reassess the growth figures every month at product office hours as usual.
  1. Growth
  • 500K tweet views (10% conversion on any app)
    • Engage a united team of yappers, everyone ups their yap score every week.
    • Weekly events that break the mold.
  • Integrations into more chains, de-risk Lifi, prudent swappers if they arise.
  • 15% MOM growth from smol/mini apps
  • Funnel optimizing (session > connections > volume)
  1. Innovation
  • Launch 2-3 smol/mini apps, beginning the “suite of shapeshift”
  • Mock and ship the multi-chain multi-wallet sweeper.
  • First AI agent prototype and infrastructure.
  1. Community Engagement
  • 25% DAO participation rate on votes
  • Maintain Discord members (14k) but more importantly engage active threads in general, product, polish, memes, community tweets, and praise. 20-100 active community members is our target.
  • Revive the subreddit? Get some sort of regular subreddit AMA’s or reddit native community engagement going.
  • 300K Twitter followers to back up those views goals

Onwards, innovation awaits.

2 Likes
2

Interesting direction, one in which we can be where the puck will (likely) be instead of “just” (it’s still a lot work) aggregating. And the immediate goals for the OG interface are exciting, I think we can reach a lot of the former ShapeShift customers with this. I generally agree with these moves!

In the later phases, are we still going to maintain a fully featured version of the wallet/app but “AI-free” (not just the OG interface)?

There are definitely users who are either anti-AI or simply not interested in AI and would potentially see these features as an added risk for their portfolio, even if they aren’t forced to use these features (yet). Potentially losing customers this way is a risk not mentioned here, but we could mitigate it this way?

You can witness that sentiment in the search engine sector for example, most of them see users build extensions to hide/disable AI generated search results. It’s certainly a sentiment that will exist for some time, like with all disruptive technologies.

2 Likes
3

Thanks FB!!!

Totally hope so!

Probably, the hope is that these are a stacked bar chart of revenues. That being said i can see a future where as transactions and onchain interactions condense to natural language we just ask "but why can’t the agent do that though? "

It’s bold but AI is just going to augment our lives in so many ways it’s silly not to take a stab. There will be anti AI people always and we are, absolutely, very early.

3 Likes
4

(im anti ai heh)
but its getting better.
i would try anything tho. (often to my detriment) … Usually to my detriment!
(but also to my benefit)

anyhow. if we can make it something like…
this amount is seperate and the ai can play with this… that would be interesting in one wallet.

(or i would just use a seperate wallet to have ai active, while its OFF in others)

Starting small ofc is the way.
would we need to find more devs for this? (ai focus?)

thanks for thiinking about this.

2 Likes
5

I think this is a great foundation and a cool direction. I think it will push our team to think outside of what we might most comfortable doing. But it will require our team to be united going down this path if we decided to trudge forward.

I am excited about this future, and cant wait to see how we can make it better as we work on this as a team and community.

1 Like
6

Yep, at a certain point once we get a POC out the door and derisk all the unknowns. We will all have to ask ourselves, how do we support this? what impact do we make as individuals? And then collectively where do workstreams fit.

For FOX tokenholders it’s simpler at least: make a thriving business model work since it uses the same/similar infra under the hood.

7

Brilliant write-up @0xFBL, I like where your head is at.

This feels directionally correct, and we are In the right place to build out the new era of ShapeShift.

2 Likes
8

Hell yeah, new era time.

1 Like