[giarc]

As others in the tech scene have said, it might be hard to find the real valuable GPTs since the barrier to build is so low. At least with the app scene, learning Swift/Javascript/ObjC created a barrier, with ChatGPT store anyone and everyone can create a GPT.

[mrtksn]

It's actually not that low. GPTs allow external functions and the moment you want to do anything interesting you actually need to be able to build it.

For example, if your GPT is going to design and host a webpage for you then you will need to be able to create a service that can talk with that gpt, take its input and deploy it.

Other GPTs that don't require a hard skill might also turn not to be trivial due to prompt engineering. An analogy would be the case of some people being very successful in literary work or on social media and others not so much.

[throwup238]

From the announcement:

> It’s been two months since we announced GPTs, and users have already created over 3 million custom versions of ChatGPT.

That's a really low barrier, especially considering that only paying subscribers have had access

[falcor84]

Your actually don't need a subscription to create a custom GPT, but rather just a regular dev account, and can play with this almost for free, just paying for individual API calls. I created probably 5 GPTs, just fooling around, and only paid a couple of bucks. I'm probably counted in their stats, even though I have no intention of shipping them.

[mrtksn]

IMHO "barrier to entry" implies success and not mere existence. It's like barrier to entry into singing, everyone can sing but the barrier to entry into singing career is actually quite high. Very few things have true barrier to entry and its mostly about purchasing an equipment or license.

[throwup238]

That's only if there's a curator/gate keeper like radio stations or major record labels and their marketing budgets (continuing your analogy). Soundcloud has zero barrier to entry so it becomes too flooded with mediocre work to be very useful unless you already know your destination or there is a powerful ranking/discovery feature.

The music industry existed long before the internet came on the scene so there are many avenues for distribution and discovery. GPTs have been sharable for a few months now and the market looks abysmal with every website/catalog/database of custom GPTs being flooded with mediocre prompt engineering.

Hopefully the official GPT store overcomes that but I'm not very confident - one of the top GPTs listed is Grimoire and a few others and I don't have a high opinion of it after trying to use it over the last few weeks.

[swyx]

> Soundcloud has zero barrier to entry so it becomes too flooded with mediocre work to be very useful unless you already know your destination or there is a powerful ranking/discovery feature.

i mean. wasnt billie eilish discovered on soundcloud? how did that happen? anecdotally feel like somehow soundcloud has solved discovery in a way that is opaque to me as a casual

[NikolaNovak]

We've just gone a lovely circle :)

The "barrier to build ", as used by the initial post and in this context, is "barrier to put something in front of eyes of the customer".

With that barrier being low, it's supremely hard for customers to pick wheat from the chaff. Who's gonna bother going through 3 million offerings and figure out what's useful?

[mrtksn]

I guess the discovery will be similar to discovery of viral tweets and YouTube videos. That is algorithmically and virally.

[robbomacrae]

And also the functionality scope and incentives are pretty poor right now.

For instance I'm trying to work out if I should make a GPT of Summer AI, well I would have to ask the user their location (like Alltrails GPT does) which is a poor experience and I'd have to forgo the usual subscription (which doesn't even cover server costs) and take whatever their "usage" based comp is.

I suspect in a few months when they add options for fetching more data, more interoperability with other apps and services, and if they add better monetization, then the GPT store will get some really interesting GPT's added.

[achrono]

That (fallacious) argument applies to nearly every other "tech" platform, from YouTube to Hacker News.

The barrier being low is not the problem -- the question is how high does the ceiling go. Is this thing capable only of preconfigured prompts based on a customized data upload, or something vastly more?

[throwup238]

Yeah and Youtube is absolutely swimming in garbage.

Without dang, Hacker News would be too.

[whitten]

Could you elaborate ? What is dang ?

[fooker]

GPT 6 bakported to the past to combat GPT 4 generated spam and GPT 5 guided scam.

[selfhoster11]

The main moderator of Hacker News comments.

[boringg]

Amen. This place would be unrecognizable.

[MattDaEskimo]

Agreed, to an extent.

These platforms have powerful, refined algorithms to handle the swarm of submissions.

So far there's a single metric of initiated conversations for rankings in the store.

The Plugin Store had no metrics, and was ranked alphabetically, resulting in plugins named "AAAA {plugin name}".

So, historically OpenAI hasn't implement any sort of algorithms