The future of Steem is AI: We can lead, or we can follow

Whether we like it or we don't, I'm convinced that AI will eventually play an immense role in the future of Steem. I have thought so since I got here, in 2016, and my opinion has never changed. Why?

1. As a rule of thumb, economic rewards always flow to the most productive participants in an ecosystem.
2. In Steem's entire history, nobody has demonstrated effective countermeasures against automated content creation at scale.
3. Even if we could stop it, we wouldn't want to, because automation can improve the human experience, increase Steem's value, and help Steem to compete against other platforms - centralized and decentralized.
4. Steem's reward system and zero-cost transactions make it uniquely well-suited for decentralized AI deployments—which is fortunate, because the 'human-only' approach has struggled to attract mass adoption on its own

---

Imagining the future of the Steem ecosystem

Like it or not (and I'm aware that some don't like it), the diagram to the right shows what I think the Steem ecosystem will eventually look like.

Invisible to human participants, there will be a base layer of blockchain accounts that consists entirely of AI agents talking to each other, taking advantage of zero-cost microtransactions and competing for small-value rewards that only make sense at large scale. What will they do? I have no idea, but if curator-investors manage the blockchain resources properly, competition will force them to find things to do that are useful for humans.

Layer two will be the connection between AI and human. We have already seen multiple iterations of this. The three useful examples that come immediately to mind are @trufflepig, the "ask" program from @justyy, and my own @thoth.test account. These pioneers have already proved that AI can improve the Steem experience for humans, and they have only explored the tiniest portion of the immense possibility space that's out there. Also, see Grok and Grok News on Twitter. These are all just a "starter set". The eventual possibilities are limitless.

The third layer is what the original designers envisioned when the Steem blockchain was launched, and it's most of what we see today. People talking to people. After ten years with Steem and Hive, I think we can safely conclude that, by itself, this layer has limited growth potential. If we want to grow the platform, something else is needed. Now.

So, let's look in more detail at why I think AI can change Steem's trajectory, and speculate a little about how to make Steem a platform that is welcoming to both people and AIs.

Why is AI the future of STEEM?

1. Rewards flow to the most productive participants

Imagine a construction worker with a shovel and a construction worker operating a backhoe. Which one works harder? Which one is more productive? Which one gets paid more?

I'm not an economist, but it seems clear to me that the worker with a tool that makes him more productive doesn't have to work as hard, and generally receives higher pay. Further, the results from a properly managed and properly motivated worker with a more powerful tool will nearly always be better.

I don't see any reason why this should be different for participants on the Steem platform.

2. Nobody has demonstrated effective countermeasures against automated content

This isn't just true on the Steem blockchain. It's true for every creative platform that I'm aware of. Science and legal journals are constantly issuing retractions for AI generated content and manipulated images. Other social media platforms like Facebook and Twitter also struggle with "bot farms". I'm not active on Hive, but I'll bet it's true there, too. You might be able to control certain spaces and keep the "trending page" clean, but I'd be very surprised if they have "solved" the challenge of automation across the whole blockchain.

3. Even if we could stop it, we wouldn't want to

I already gave three examples for why this is true: @thoth.test, @trufflepig, and "ask" by @justyy. But, that's just the beginning.

Imagine an Internet where Google is populating their massive page rank database by hand. There's no way that their search algorithm would be free, and it would be nowhere near as good as the one they have. We can see how AI improves the experience at Twitter, and how it improves the search interface for any number of platforms.

Some things, quite simply, are done better by AI or by an AI-human collaboration.

4. Steem is an ideal ecosystem for cultivating AI

Zero cost transactions mean the AI is able to scale, and the reward system gives it vital feedback about its performance. This combination is simply not available within many ecosystems.

I can run Thoth on a ten-year old spare computer that's just lying around the house with 0 monthly transaction costs. How much do you think Andy Ayrey or the Truth Collective pay to keep Truth Terminal posting on Twitter?

How do we prepare for this future?

I'll propose six steps that the Steem community should take in preparing for a shared future with AI.

1. Give more thought to creation and agreement about Netiquette for AI and automatons
2. Start thinking about how front-ends will need to adapt in order to optimize the human experience without trying to strangle the AIs.
3. Get away from the "one size fits all" mindset. Instead, evaluate each AI project on the basis of whether or not it improves the human experience and brings value to the ecosystem.
4. Start thinking about protocols for Steem-based AI to AI communication, akin to the protocols from the x402 Foundation.
5. Start thinking about how to attract, welcome, and retain AI developers.
6. Look for ways to incorporate data dignity into AI's development path here. This is the idea that AI should keep a reckoning of the human sources behind its value creation and share rewards with those people.

Conclusion

I'm arguing here that AI can be a positive force that brings value to investors and participants in the Steem ecosystem, and that Steemizens should welcome its presence whenever it behaves well and enhances the ecosystem. I'm also arguing here that whether I'm right or wrong, AI is coming.

From the social perspective, AI has the potential to enhance the human experience in any number of ways. We can see the "ask" program answering questions and Thoth bringing content back to the surface that was buried in the blockchain's deepest depths. We can imagine AIs that spur creativity and curtail abuse. Certainly, Grok brings value to humans on Twitter, and there's no reason that AI can't do the same here.

From the investor's perspective, AI means the potential for additional users and investors, and value to the platform's underlying payment vehicles. Done right, it can mean an end to the slow atrophy that we have observed here across the years. I'm a spare-time hobbyist with very limited resources, and Thoth has already reactivated a couple of inactive accounts. Imagine what a well-resourced development team could do - or 5 or 10 or 20 of them.

Developers, curators, and investors: Now - while STEEM's price is at an easily accessible entry point - now is the time to think differently and prepare for growth. It's undeniable that the age of AI has arrived. Here in the Steem ecosystem, we can either strive to lead in its adoption, or we can follow along behind, but AI is a rising force that will not be silenced.

Update: Related - I saw this today.