LISTENING MATERIAL: As a continuation from my previous post on Keith Jarrett about the best-selling piano solo album of all time... Well, it's not necessarily the best just because it's the most popular. With that in mind, listen to this while reading, largely considered to be the best of his improvised work:

(removed cause of thumbnail glitch)

Now, there was a long time in the past where there was a lot of love, and a lot of hate, for curation groups such as STEMsocial, OCD, and all of those in between.

I get the vibe that this hate has subsided, and it's largely accepted, and that makes me happy as I am obviously on the PRO side of this.

I can, of course, see the negatives - particularly about voting trails being idle and getting rewarded for auto-curation, as in, doing nothing. Among many other things.

However, looking forward, we have a greater problem: AI.

Yeah, another AI post. Woo.

Recently, I've been using zerogpt.com to scan various posts to find how much of the text is likely generated by AI. If I get a positive, I take it to ChatGPT itself and directly ask. GPT typically confirms it was written by a bot, and by itself specifically.

Now, it's true that it's a little early to consider this totally infallible. But at the same time, nobody wants to risk wasting their stake on the blockchain on what could potentially be bot material.

In terms of effort and creativity, it is no different from your cookie-cutter plagiarism. Sure, the words can't, in the exact order, be found anywhere else. But at the same time, it is not your blood, sweat and tears that went into your Entrepreneurial push to earn for creating content.

It is also difficult to argue with the accused. They can simply say Zerogpt is wrong. ChatGPT is wrong. We can't prove beyond reasonable doubt. They can say 'I just used it to proofread' or 'fix my english'. There's a world of excuses, and a lot of workarounds such as making a few changes in each paragraph of the wording to trick the detection tools.

For the average reader, perhaps this doesn't matter. Youtube for example is about 80% bot content now, with videos generated by AI, bots doing the voiceovers, and bots creating the text. Everyone has become so accustomed to those stock voices on Tiktok that it has become the norm.

As long as people get the content they were looking for, most simply do not care, and the frauds that create these youtube accounts (sometimes hundreds of accounts posting the same basic stuff, by the same single person, jumbled up by AI amounting to millions of combined views for barely any effort) make bank.

The same applies across the board. Here on Hive, people might read through blogs, drop a comment, and get a basic reply, and their interactions are thus satisfied, despite the dishonesty of the whole thing. I mean, just look at my terrible AI pics. Nobody cares, really, that I didn't hand-paint them myself. Can you even tell which one ***isn't *** made by AI?

Follow this trajectory long-term, and Hive becomes 90% AI-generated content, which will improve over the months and years to the point that it will establish personalities and humour to engage people even more with no effort or true individuality.

Eventually we will start to notice a common theme from all the authors. The same writing styles, the same phrasing, the same structure, the same frequency and schedule. We will stop seeing updates to life and projects, but simply clickbait about news or fun facts.

And, without a genuine personality to follow, what is the point in using Hive at all other than money? People don't follow Taylor Swift for her amazing compositional masterpieces. They follow her because she's constantly breaking up with boyfriends and writing songs about them while looking pretty. People follow personalities they relate to.

The Case for Curation groups

This is where professional curation teams come in. These folk are not simply consuming content, a decent group have a trained, critical eye. They scan for details such as plagiarism, false information, copyright infringement, engagement with the greater community, and AI.

It is, as far as I'm aware, the only protection against such a dull, omnipresent bot platform.

Not only are they simply policing the fraudsters - by means of not supporting them - but they might also give greater support to trusted users, those with a larger, earned following, or perhaps just those who write with personality.

The greater influence a good curation team has on the platform, the further in time we push back the inevitable bot coup. Even though AI is easy to make, it will be less and less worthwhile if all the confirmed genuine writers are taking all the rewards, and the AI basement dwellers are earning 3-4 cents per week.

What should we do?

I think it is up to each group to decide what to do about this, but it should certainly be a priority discussion.

For STEMsocial, I am in discussion about erring on the side of caution with zero tolerance:

• If it gets flagged as AI, No vote.
• If the user proves it is not AI, still no vote
• If a second time, perhaps we will avoid curating them indefinitely

Why do I suggest not voting even if it is proven not to be AI?

Simply, if you are creating content that is mistaken for AI, then you're not really creating anything worthwhile. If you are using it as a correction tool, then you are taking the laziest possible route to improve your own writing - and art - style.

There is zero need for chatGPT or anything similar on this platform. People have been writing for thousands of years without it. There is no reason that it is suddenly an essential tool for us here, either.

I am not sure if this is the best way forward, but that's what feels right to me at the moment.

In the meantime, I commend every other initiative that exists to filter out the garbage and promote the good stuff. Without that, Hive would be a far more desolate place.

So what do y'all think?