Oddbean

First iOS Nostr App to support #DVM You can now generate AI image right inside the Current App and share it on post with just 1-click. Try it out now. It is out on testflight..! If you want to experience the hype of DVM.. Go for it. https://i.current.fyi/current/app/iamgegen.png"

Great job! nostr:nevent1qqstp6wxhq2qqmftwxmpzvsnnvtqpc2rqgm6hwy6gvc933jd5ssfsrcpr4mhxue69uhkummnw3ez6vp39e3x7mr59ehkyum9wfmx2u30qgsvwp3ue4lf4hqdm49h034l40luswvmg83ymcaxdzn2kchduty240grqsqqqqqp6qykxk

This is really cool to see. Hadn't thought if this use case before. A really good use of image generation DVMs! Congrats to the @current team on leading the charge and innovating! Now every developer can improve @current app without touching its codebase. This is why open networks win. nostr:nevent1qqstp6wxhq2qqmftwxmpzvsnnvtqpc2rqgm6hwy6gvc933jd5ssfsrcpp4mhxue69uhkummn9ekx7mqzyrrsv0xd06ddcrwafdmud0atllyrnx6pufx78fng564k9m0zez4t6qcyqqqqqqg7ncqhu

Thanks @PABLOF7z for your support and leading DVM specs. The plan is to get it right for 1 DVM agent/bot and then open it up to start to list other DVM bots that can satisfy the DVM request. We will even stop the Current bot once we have enough agents/bots from the community.. Network effect starts right there…!

LFG! This is the way. Let me know if you need help; I built on Vendata.io a generic approach to the way I see this working. I’m going to be demoing soon a narrower implementation that leverages Content Discovery DVMs in a way that I think will be very relatable to many clients

For example, on Highlighter this is how I'm allowing users to choose a DVM for "Suggested Content" What's on the dropdown is results from a NIP-89 query for #k 65008 (nostr content discovery); when the user chooses one of those it creates a 65008 event `p` tagging that DVM. What's behind the dropdown is the rendered result of the tags the DVM sent back. If other DVMs respond to the query I can show a badge on top so the user can cycle through different DVM's results of content discovery. Right now both results are from different DVMs I run, but when Semi finishes his stuff that will be another DVM, and anyone can extend the functionality of Highlighter (and any other nostr client) without touching it's codebase https://image.nostr.build/1f059d3eb73d7b7b9aea799a7f89ee5ef0a7f80d47c7b18acda3d31668cc92ad.png

@PABLOF7z responded with the answer. At times he meses up the event kinds.. :) Image Generation event kind is 65005. https://vendata.io/dvms shows all the DVMs.. find the kind 31990 event that says it supports 65005. We are furiously working on to expanding the support to list all 31990 with support for 65xxx I will share with you some docs next week to show the architecture and workflow. {"created_at":1693518654,"content":"{\"name\":\"Love AI Image Gen\",\"image\":\"https://i.current.fyi/help/Screenshot+2023-05-10+at+10.18.33+PM.png\",\"about\":\"Generate image\",\"nip90Params\":{\"relay\":{\"required\":true,\"values\":[\"wc1.current.ninja\"]}}}","tags":[["k","65005"],["d","pomi8fh9bxpntgfy"]],"kind":31990,"pubkey":"c70735fa4b01f77f953883a6e671982e31bd7d906b2b6111a6f518555bed1b1a","id":"82ac9efbff7822cc7375858d4c4b6c398e75331d5df637bb5c6abab68da350f7","sig":"e27f5f56cd592a6af63d8168a16bc64bbbacdef0823f6ec43e6f58d950e79d66d12f5008eba559a728b8d1f195436d49ce5f2c29d9f9d9c15a0708942f16236e"}

Hi @current I just tested the DVM generation in your app. Very cool! One question, I see that you are not listing the results of the other DVMs that are responding, but rather fetching the result of one of the DVMs that respond Are you planning on adding that and allowing users to choose? That'd be rad! I know some DVM developers want to provide specialized DVMs that are trained with particular styles; would be very cool to be able to support that in your app 😀 https://image.nostr.build/d00de1de55c5dfd6eb26526fda197c353b3d1786085aca497155032c687d5761.png