in general its pulling from training data so the first numbers picked are always going to be pretty similar. The AI is worried about not providing variety (for example picking 7 for 10 rounds) so it says to check the chat for context on round 2.

on round 3+ it feels like it solved the problem and doesn't need to evaluate anymore and it inadvertently uses more effort to create something less random

I did something similar.

When I started, AWS was in its infancy and I was just some guy working on a special project.

Now that same account is bound into an AWS Organization.

AWS Changed. My company changed. the policies change out from under you.

if your movie collection is big enough that might be really useful for language learning. Create your own frequency lists and common phrases. I would be curious how it stacks up against the written word.

That's exactly what I'm using it for! And my movie collection is pretty big. I have about 1000 movies. (Many are translated, but also many are not.)

payment would need a delay too. Pay $10 and then wait a week or so for the payment to clear without it being reversed. Hopefully that stops the card stealers from dumping as much as possible before getting booted.

Could we just add complex and varied captcha to the comment & posting forms?

That's not a bad idea, sending mail could simply be an authorization for a $1 or $10 charge. And if the receiver said the message was unwanted, then the charge would go through.

There's just the pesky problem of incentives on the other side of the coin - who gets the $? The spammee? But there would be enshitification issues like:

1. Those who are incentivized to take as big a cut as possible.

2. Those who would put it in their EULA that you must accept their spam and not chargeback or else you lose access to something you value like their services (EULA Ransom... not much different to today "accept our EULA or lose access to what you've already paid for!")

I'm sure there are many other perverse incentives which would creep in..

His backup plan prior to the event had large obvious issues.

His backup plan after the fact seems suspicious as well because he is making it much harder than it has to be.

Between that and a glance at the home page, it feels like someone doing AI vibe work who is not comfortable in the space they are working.

Who is the intended audience? Other vibe coders? I just think its weird that given his backup solution, he likely asked the AI to create it . whatever hot-wash he did for this event was invalidated.

The registrars I use are limited because of KYC. the network of registrars under dropcatch all require KYC. Many of the auction sites require KYC.

The problem is they hold faces+IDs indefinitely and the ones I needed have breach issues along with indicators of top of the line 2005 era security practices.

So for me, its less about the KYC and more about their inability to KYC responsibly.

I thought "surely they wouldn't ...." The issues in the article are more blatant. You were right and caught it extremely quickly.

please make the translated text a little bigger. On a phone at that speed, it can be hard to read.

You need an optional login. Forcing me to login before I know who or what you are is a deal breaker. Especially if you come here looking for an opinion. My opinion is free. My PII is not as free.

Love Zen.

I am not sure it fixed my tab problem, but it improved tab management and overall efficiency.

Omnivore helped a bunch before they shut down. I need to see if any good apps exist that are similar.