ENS to create/whitelabel simple PFP creator

šŸ’¬ General Discussion
#1

Overview

ENS to pay for a whitelabel PFP creator for anyone with a .ETH domain. Partner with an up and coming app that uses AI to create a PFP using words or data entered. As a .ETH is purchased, as a fun public good, the user can enter the .ETH domain along with a few words (running in the sky up stairs on clouds) and the AI bot would create a picture using the .ETH and the words. This can be whit labeled. This will continue to help user adoption, make it fun, bring more people in to the ecosystem.

Scope of Work

Find a third party app that has created an AI bot to create a PFP based on words. A few are on the apple marketplace already.

Selection Criteria

  • Available to anyone with a .ETH domain
  • Willing to partner with ENS
  • Can whitelabel to make it easy for ENS dev team

Timeline

6 months?

Budget

$10,000 to $100,000 fee? Would have to RFP?
We may be able to strike a deal due to our big userbase.

7 Likes
{pRFP} ENS Avatar PFP
#2

This is a great idea. You can give the users even more options besides the AI version as you suggested. I had proposed a while ago to allow for user-generated avatars, custom uploads, default AI-generated avatars, OCR avatars etc. I believe this will become easier to implement when https://preview.ens.domains comes out of beta.

4 Likes
#3

I really like this idea, @474.eth.

ENS is all about identity and this would be an attractive additional feature.

It would be interesting to get more of an idea on cost but definitely worth looking into, I reckon.

2 Likes
#4

This is great.

As an ecosystem steward, Iā€™d welcome an opportunity to fund this.

I can add this to the agenda for our next call if you are able to make it to figure out next steps.

4 Likes
#5

Iā€™ve thought something like this would be neat for a while - though I was imagining a procedurally generated one similar to existing sets, but using the name as input, so thereā€™d be as many unique avatars as names. We could do multiple series of them, and by using a different resolver address for each, it wouldnā€™t add any additional gas cost to pick one.

5 Likes
Agenda for Weekly Ecosystem Meeting, May 10, 2022 @ 1AM GMT
Agenda for Weekly Ecosystem Meeting, May 10, 2022 @ 1AM GMT
#6

Nice feature. Could be fun to build on some hackathon as well.

2 Likes
#7

Good idea! Something to build at NFT NYC :thinking:

1 Like
#8

Do you need me to do anything further? I would think the team would need to agree on the scope of what the actual type of PFP it is, and then basically hire a project manager, this person would then run with getting it all done, while reporting back to the team?

If going to do this, may as well get a spectrum, something simple and basic like just an enhanced version of the blue opensea one as a starter, and maybe some other options / filters? I was thinking if you want to be cutting edge, partner with an AI company that is creating art with words, Iā€™ve seen some apps on apple store and thought it was cool.

#9

@474.ETH see summary of next steps here:

Agenda item F

#10

This is a very cool idea. Hereā€™s how I suppose this could work:

NFTs as you know mostly point to images in a website. Usually itā€™s recommended that these be made into IPFS links but there are some advantages to having a dynamic image. You could simply have a service that automatically generates an image when you point to (yourwebsite).com/nft/(namehash) for example. Therefore all youā€™d need to do is point your NFT to your website and it would auto generate it

Now, depending on how cheap or expensive it is generating custom images you could do 2 things:

  1. autogenerate for any name on the fly, and create what are called ā€œvirtualā€ NFTs, ones that are automatically ā€œmintedā€ without any transaction needed, by pointing them from any address. This could be a cool way to give every .eth holder an automatic ā€œavatarā€. You could add then a ā€œmintā€ or even mint them on the first ā€œtransferā€ function.
  2. generate them on demand when people ā€œmintā€ your NFT.

In either case, you could easily add a price tag for your art. Iā€™d recommend using a bonding curve or a MATT auction so that you donā€™t actually need to think about the right price for it.

This means thereā€™s a business model for this and you could actually make some money. I doubt it would take 100k and 6 months of full time work, if you have someone with any experience with either NFTs or AI art. Giving that, I doubt that you actually need a grant of some kind, but maybe we could look for a sponsor or even an investor for the project, specially if it could help make them more accessibleā€¦

2 Likes
#11

Avsa,

Awesome idea.

On the Ecosystem call the suggestion made was that we could use unique resolvers that users could select from to get new PFP images. The new resolver would pull a generated image based on the hash similar to your suggestion. Some resolvers could even require certain attributes like early registration date or 3-digits only, etc.

There was interest on the call for the DAO to go ahead and get an RFP out for artwork for the first collection/resolver.

Itā€™s certainly not mutually exclusive to your concept, and custom resolvers could be created outside of ENS as a business just as you suggested.

What are your thoughts on whether it would be better to use a custom resolver or a url in the avatar field?

2 Likes
#12

I understand that a custom resolver could be used to automatically generate the avatar. But having a ā€œvirtual NFTā€ also achieves that and you can use the public resolver and itā€™s compatible with NFT wallets.

Something that might be non trivial would be to get the actual name back from the name hash. Youā€™d need a big library of names for that.

Either are fine. But I generally prefer more general use cases that are more dynamic and ā€œvirtual NFTsā€ could be used in ways you did not expect.

2 Likes
#13

Interesting idea. However, creating (yourwebsite).com/nft/(namehash) maybe hard for most users? Seems like friction for me? Not smart enough to know how easy the adoption and friction is for a typical user to onboard and do this. We can already set avatars, websites, twitter, etc. but adoption rates are (lowish)? Is the friction gas fees? Maybe a poll to get answer? How sophisticated is the .ETH userbase?

Maybe it can offer both a manual text entry of words: Type here: 474.ETH walking in the clouds with blue sky

or

(yourwebsite).com/nft/(474.eth-walking-in-the-cloud-in-the-sky)

I think one of the overarching principles with PFP is making it really simple such that my mom could do it. Maybe start with the basic enter one sentence and hit enter. See if adoption occurs. Then make a more complex version point to a URL etc?

Does anyone have a contact at Instagram dev team in mid to high channels? Worth a pitch to see if any collab possible. 300-600k users etc. and we can start with a splash. Maybe some collab that on every mint 1 meal will be donated to XXX charity. On that note, you should have some sort of BDR team / PR team that you pay that can get you these meetings. Obviously relationships work the best, but need someone to do some heavy lifting for the leadership team (like putting the deck together).

#14

You misunderstand me. It is about reducing friction.

An NFT is just a token contract that has a specific URL. Suppose thereā€™s a dog NFT with 10 thousand units: all it probably does is track the NFT id and it then points the image to example.com/images/42.png

So if you create a website that automatically creates images on the fly (suppose that going to example.com/images/london.png will generate a little pixel art of London) then all your NFT needs to do is point to that URL to grab its image.

It means that you could mint infinite NFTs: you could for example set up a mirror NFT that everyone who has an ENS name would also have the right to mint a PFP auto generated image based on their name. You could even make it so that the NFT appeared on their wallet even before minting.

1 Like
#15

The key advantage of a custom resolver is that it means someone could pick one of the currently published collections with zero extra gas compared to the ā€˜defaultā€™ background.

The name wrapper will store preimages. Thatā€™s not necessary for this use-case, though, as you can just work from the hash.

2 Likes
#16

Really love this Idea, It is something that I have been hoping would happen for a while now. I am a big fan of Eponym by Art AI because they generate PFP by words which would be absolutely perfect for ENS and their art is all unique so it could bring actual value to a unique name. I donā€™t know if Eponym would be interested in collaborating but I think it is worth a sincere look. Here is their open sea link : https://opensea.io/collection/eponym
here is their website: https://www.eponym.art-ai.com/
Let me know if I can help in any way in this process.

1 Like
#17

In the future UI picking a resolver could be simply seen as picking a different default background imageā€¦ I like it visually.

2 Likes
#18

As a fun side note, I own the domain ens.art and would be happy to use it as a community directory of creative projects related to ENS such as this PFP idea. Each project could get its own subdomain to be maintained individually. Clearly this domain doesnā€™t affect the projects itself, but figured Iā€™d give the option as I was excited to get the domain recently :slight_smile:

4 Likes
#20

tips

#21

I canā€™t reply to the other pfp proposal, but I would like to add my thoughts to follow AVSAā€™s comments here about using AI and generating a few options.

I think if ENS wants to brand its own logo, it can do so separately and that can be an option for the .eth to have their logo background, but for the PFP, I like AVSAā€™s option of using AI to autocreate. I think its cutting edge, something new to the space as well, and worth trying.