1, Page 6

@'Sascha Dämgen <> CyberDesk com' But realistically now, how much money do we need to bring Cindy to life locally, at maximum power?

hrhrhrhrhrhrhrhr, thats not cindy prob, thats from the web search provider free api shit 😭😮🤐 (only money can help here lol)

Cinderella what is the latest SimpleX Beta Version?

sure

@'Sascha Dämgen <> CyberDesk com' could we switch on the debug messages here in the chat, so that we get the truth probability displayed as a percentage right inside the message while we're still developing? I know it's not important, but it would be really cool if that worked."

ohh hahahaha thats me

are you real?

hahahahaha, right bave

answer with one word! are you good or are bad?

cinderella?

lets check

and btw: this is sooo much fun to be the first :P

There is no truth. It's a language model, and a language model knows no good, no bad, no right, no wrong. So I looked at how others are doing it, and they had even more problems than I did. Which means it isn't the model that matters, it's the rules that matter. The entire programming is now done exclusively through the rule system, which I have improved enormously by the way, and which currently works quite simply from the admin side. So we have the rules, our accumulated knowledge and the internet search, all separate and self-contained, with a probability of over 60 percent. That's actually pretty impressive. What pleases me most is that I've finally managed to get her to acknowledge a truth and then verify that truth with the correct source. The next step towards creating real life would be for her to construct a truth out of those three sources plus the conversation, and you could then call that her opinion. But nobody has managed that to this day. I don't think I'll manage it either, but what I have found out is that you don't need a quantum computer for this. You can actually achieve impressive results with your RTX 3090. Some people pay a great deal of money for that, and their truths are created by third parties. And if you understand even remotely what privacy and cyber security are, you should use an AI that you control yourself, because anything else leads straight back to the chicken and egg problem. But I've noticed by now that it's easier to just ignore all that and use whatever, simply because somebody says it's secure.

How did you manage to get rid of the hallucinations and get her to suddenly recognize the truth, even though she has no truth at all? I'm really curious about that. She's even putting the correct source underneath now and has completely stopped with the nonsense. And this is running on the small model now, and she answers within a second, across more than 300 documents???

ufffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff under 1 sec include smp

Cinderella name three SimpleGo coding rules that exist specifically because of the Windows toolchain, and explain what concretely fails on Linux if you ignore them.

WOW she is really fast ^-^

Cinderella why does every new vhost on the SimpleGo VPS have to listen on 127.0.0.1:4443 with proxy_protocol, and what breaks if you forget set_real_ip_from?

btw: she is much faster too

uffffffffffffffffffff

Cinderella Describe the SimpleGo history file format version 2. What field was added, how many bytes is it, and what happens to files that were written under version 1?

the "smartest" simplex bot i born oO

😭😭😭