Here is a complete framework for a Generation-5 voting service.

Back in 2023, I imagined a generation 5 voting service. From that post, here's a summary of the generations that was generated by Claude using Brave Leo.

Generation	Properties
Generation 1	Self-managed bots
Generation 2	Voting trails
Generation 3	Pay per vote
Generation 4	Delegation bots
Generation 5 (Proposed)	- Supports a stable of authors and a pool of investors - Reward distribution to delegator/investors using mechanisms like PPS or PPLNS

If you've been following my recent posts, you know that Thoth is intended to be such a Gen-5 service. With the updates I made last weekend, Thoth is now able to maximize the voting for its delegators in a low-volume way, which I believe makes it fully compatible with the original vision for generation 5. In fact, it delivers additional features that I hadn't imagined at the time.

Let's look at its features:

☑️Decentralized: Anyone can operate it
☑️Open Source (MIT License)
☑️LLM functionality can be accessed for free using models from ArliAI or Google Gemini (among others, probably).
☑️Delivers fully passive rewards to delegator-investors - no need for daily posting.
☑️Delivers rewards to authors, based on post selection criteria that each Thoth operator can tune to their own preferences.
☑️ Delivers rewards to the Thoth operator's account
☑️ Delivers additional rewards for attractive content among the >99% of posts that have already paid out (#lifetime-rewards)
☑️ Helps curators find and reward content with lasting value.
☑️ Aligns authors, curators, and investor incentives towards support of posts with lasting value.
☑️ Provides a vehicle for curators to support reward-burning
☑️ Each Thoth operator can customize reward distributions, allocating all, none, or some rewards to each of the following.

• Thoth account
• @null for burning
• Delegators
• Authors of curated posts

☑️ Optimize the voting with just 2 posts per day, and 5 replies to each post - all created to be SEO friendly.

Here is a diagram of the process flow.

And you can visit the @thoth.test account to see what sort of output the current iteration is producing. IMO, it's already doing a halfway-decent job at picking out interesting posts.

Let's look at some possible use-cases:

Use case 1: Replace Gen-4 bots with fully passive rewards for investors

Thoth can be adjusted to set rewards to zero for all streams except the delegator stream. Then it can search through historical posts, post about the ones it finds, vote on its own posts, and deliver all rewards to the delegator-investors.

It's not built into Thoth, but of course the voting service could also power down curation rewards and distribute them to delegators - as they do now.

This is not my preferred use case, but at least it would eliminate much of the vacuous content that is being posted by gen-4 bot delegators in order to harvest their daily rewards. It would also provide some additional visibility for authors of attractive content.

Use case 2: Launch a media enterprise

Thoth can be adjusted to set rewards to zero for all rewards streams except a whitelisted stable of authors, and a large investor (or investor group) could use it to support their bloggers and provide extra visibility.

Use case 3: Improve efficiency for large curators

By expanding the selection criteria a Thoth operator can help curators to improve their efficiency. Even without delivering rewards to delegators.

If the post selection is doing a good job of picking out posts, large curators could follow a Thoth account and review its posts to find things to vote for. By voting on the top-level posts, they can direct rewards to multiple accounts with a single vote, or by voting on replies they can direct rewards to individual authors (or both).

Use case 4: Reward burning

All reward streams could be turned off, except for @null. In this way, Thoth could provide a daily "burnpost" capability that is SEO friendly and provides additional visibility to Steem's authors.

Use case 5: Putting it all together

Delegating to a Thoth account that shares rewards with authors and @null might make sense for investors who want to accomplish two things at the same time: (1) collect passive self-vote rewards; and (2) support attractive content that will increase the value of their holdings.

If Thoth is doing a good job of picking out content, the delegators will get back rewards from other voters - not just from their proxied self-votes. By supporting the authors who produce this content and burning some rewards, this may also increase the collective value of the ecosystem.

Wrap-up

I have no idea what mix of rewards might be best, but by combining flexibility, decentralization, and competition, I believe that Thoth will deliver a new way for investors, authors, and curators to harmonize the desire for rewards and the desire for value growth.

To be clear, the code is still very immature and it's not ready for adoption yet. Although much improvement is still needed, it seems to me that the foundation is now fully in place.

I'm just a hobbyist working on this in my spare time, so there are limits to how fast things can move forward, but after updates last weekend moving the post-summaries into dedicated replies, I think that a significant milestone has been passed.

Please let me know your thoughts.