I have read the article and to be honest I did not notice for a single time that the article could be written by an AI. If such posts can be generated from AI, I think we have some in our forum. The article was correctly written and is far above grade in term of quality from most of our forum users.
Some days ago, I think someone had reported user who were posting randomly. May be they are the creation of GPT-3.

That might cause big trouble in our community (or even other forums) because we aren't able to determine the characteristic of posting on this AI bot since the bot can able to generate text that relevant to the topic. Sound like prediction answer.

But I noticed that there are some awkward words that the AI bot has been used. And the creator was right, probably we can determine in this way.


If someone reply or making a post something like this conscious to unconscious probably they're bot or spammer.

With the fast developments in IT and machine learning, bots will be upgrade to smarter AI-bots and I believe that they will be able to wisely enough to create posts and join discussions. It is the thing and change to come to our forum, obviously. You can see the same thing with Google translator. Years ago, translation products from Google translator are terrible in quality but now its quality is acceptable enough and can be used for people can not translate original documents by themselves.

Do you think we can avoid such AI-bots evasion here? I don't think so.

But to say AI-bots can damage or destroy our forum, I don't believe they can.

We have:

• Merit system: to prevent spammers, and bots (even with AI-bots) to rank up. If needed, merit system's upgrades can do better to prevent them.
• Bump score: to prevent bad effects from bump services. Even if tehy are AI-bots, they can not beat the bump score and the algorithm of bump score can be upgraded too.
• Combined effects from merit system and bump score: they can not rank up, their bump score are nearly zero and (if >0) can not increase.
• Reports from community can help: to delete spam posts, bump posts, and nuke related accounts.
• Avatar and signature disabled can help: I think AI-bots are smart enough to realize those changes and stop posting

Basically, who create bots, even AI-bots? Human, so human can beat bots.

---

I don't like Facebook but they have a feature to disable rights to share a livestream if finding something suspicious on exponential shares in short period of time. Maybe sometimes findings are false ones but it is helpful.

I lose brain cells everytime I read one of his posts lul.

Unless theymos gonna make or use AI which can detect whether a message is written by AI/not, there's nothing much we could do aside from reporting posts that doesn't make sense.

There's simpler & good horrible solution for this problem, which forcing user to solve CAPTCHA every time he made/edit message.

I have no doubt AI research is going to change things, but bitcointalk forum may also change in time (I hope).
One thing we could see is some way to prove we are humans with some kind of sybil tests, or with new way to log in the forum.
Lot of bitcoins are spent for new forum, and I think forum proof-of-person system can also be considered for future.

I for one welcome our new AI overlords. Finally I can ask "is CSW the real Satoshi" and get an honest answer. Or most likely get an irrelevant salad of Google search results and tag the fucking bot with a neutral... actually even red trust would be probably ok since I don't think that trading with a bot would be advisable at all but I digress.

I haven't seen any AI capable of having an actual opinion, let alone style or personality, and the article hasn't convinced me. But it might be able to fool a signature campaign long enough to get paid so there is definitely a niche here.

If the article posted by OP is really written by using AI, then I can only say "Houston, we have a problem". I believe that the predecessors of GPT-3 are in use and have found practical application, especially on forums and some social networks, and that we will soon get literally virtual characters in the virtual world which will very faithfully imitate real persons.

If I have well understood the whole procedure for AI to write an article, it comes down to the setup of basic parameters on the basis of which then GPT-3 writes that article. I looked at some of Manuel Araoz's tweets and here's what it looks like in practice.


https://twitter.com/maraoz/status/1284581042175836163

We already know what the answer is - https://bt.irlbtc.com/view/5124947.0



Or perhaps everyone will have to read their posts out loud on video, but keep an eye out for deep fake Tom Cruises if they forget to reset it.