[rglullis]

Paying for the service does not give the right to spam. Any spammer would be kicked out on the first report, which would make the whole operation expensive and impossible to automate. You can not generate valid credit cards automatically like you can simply generate a bot that signs up for free.

[KomoD]

You very much can generate a ton of valid cards, or use stolen ones. In fact, people have already started using Twitter Blue as a way to help with their spamming (both crypto spam, and t-shirt spam)

[rglullis]

> You very much can generate a ton of valid cards, or use stolen ones.

Then you raise your fraud detection controls and only activate the user after payment has been cleared out.

The point is that spam is not economically viable if it requires payment. It's not rocket science.

[KomoD]

> The point is that spam is not economically viable if it requires payment. It's not rocket science.

That is very much not true at all.

[rglullis]

Please be specific: how do you think it would be viable for spammers to create bots on Twitter (or any other social network) if it costs (at least) 1 dollar to set an account and if you still keep systems in place that can detect/report/suspend accounts that show bot behavior?

[jrflowers]

I agree with this. What many people do not realize with the ease with which anyone can spin up thousands of bots that absolutely sound human nowadays, spamming has never been more difficult or less rewarding.

All Twitter would have to do is invent and maintain software that reliably detects ai-generated text with zero false positives. Done and done.

Should they do that and the spammers still seem motivated, they would have to do something outlandish like employ large amounts of humans to replace the bots, which isn’t something that has ever been feasible for spammers.

People really underestimate quite a few things, like Twitter’s ability to simply invent bot-fighting software of heretofore unseen power and complexity, the difficulty spammers have accessing dollars and other resources, and how unmotivated and prone to giving up spammers tend to be.

[rglullis]

Sarcasm aside (did you use chatGPT to try to make a point here?), the argument is not that charging makes spam impossible but that it makes economically inviable.

Just assume that it costs one dollar to sign up. The "spam-detecting" algorithm doesn't need to be foolproof. It just needs to find the spammer by the 10th message to effectively make each message cost $0.10 to be sent. What type of scam/spam has such a high ROI that can justify this operation for long?