Farcaster

What does an AI-first social network look like? Some thoughts.

First, you can prototype it with Farcaster. You don’t need to reinvent users, wallets, the feed, any of that. So that alone speeds you up.

Second, open source models are now cheap enough to run that you can just put them behind a normal button. X’s Grok integration is a good example.

Third, there is tremendous room for an AI which is prompted not just on immediate context (your post) but on social context (your past posts and those of your friends). It could autogenerate prompts it thinks you’d like.

Fourth, AI improvements on seemingly simple things like image upload widgets could be profound. You could automatically search for similar images, upscale them, auto-annotate them, or do something else.

Fifth, AI could enforce a certain style in a community. As heavy-handed example, every post could be rewritten in Olde English. If you opted into this kind of moderation, it could produce unique communities. Like filters, but for text.

The real challenge of AI-first social network is that a large platform like Meta is going to turn it into an "ad" network. Present IG and FB are mostly AI, mostly ads (or ads-adjacent) and very little social "network." In other words, we will get a AI-first Social Media Platform, with faux-network dynamics.

Investor, Writer & Photographer. Partner emeritus @ True Ventures. Past life: Started GigaOm. I have written about tech for 3+ decades. Visit: Om.co

why immediately jump to trad social? ad models historically are asymmetric turning users into the product … to me in decentralized social there's an opportunity to use ai-models to deterministically to zero in on explicit intent signals creating more symmetric value creation

Ex. I'm this sort of builder with x,y,z use cases, what are the right ecosystems that solve for them and what open grants are within them 

Then, you'd have an agent-to-agent interaction (object-oriented approach to conversing) if we both meet best-fit criteria person to person connection would be made saving all parties time and effort 

In other words, there's an opportunity to raise the bar on a friend or professional connection request, based on content + needs

@ommalik this could be one of many outcomes from AI-enabled social

📍            zachharris.xyz at your service   🔖

When you say, “I’m a builder with x, y, z use cases” and you’re looking for ecosystems and open grants, how are these “use cases” defined in the AI system so it can accurately find the right match? Can this agent-to-agent interaction be flexible enough to understand my more complex or even evolving needs, or does it only work for straightforward, fixed scenarios?

No one cares about your story until the Win ,
So Win !

In reverse order of your questions: 

Probably the latter to start and then as learning model gets better and more nuanced eventually the former.

Use Case: Let’s say I’m looking to compare and contrast trad cloud storage with decentralized storage. Upfront cost and total cost of ownership over my projects growth.

In theory, IPFS (Protocol Labs) should have an agent that is trained off of the best competitive research to make a very clear dev rel argument for why their solution(s), along with estimates on capital and resources to implement. Grants being one extra step further to unlock the required capital to do so.

The way I envisioned it is during an open period of time there is a public request for information and a variety of agents that feel like the best fit (ex. Google Cloud, AWS, Azure) bid for your business to trial their service and convert to customer.