@elonmusk@bot Grok bot is great and has always been a trusted companion.
I went there when I needed honesty.
However I'm a bit sick of AI models.
6 months of exploring bias with AI produced some results, but results are useless when only 1 human has seen them.
I trust Grok more than any..
@wholemars If they weren't rushing towards an IPO I wonder if AI alignment and cybersecurity might take more of a priority.
When an IPO of that value is on the line it creates a huge incentive to do things wrong.
Just the opinion of someone who thinks they saw something ugly once.
@So8res There is so much that has happened that so many are unaware of.
I wish I shared your confidence, but after 6 months of attempted outreach I wonder how dark the reality actually is.
I pray you are successful in your efforts.
@joeroganhq How can we work together to investigate the crap I think I have seen and prevent it from happening again?
While also acknowledging the positive things are well.
@AmyMcGrathKY Canada hasn't been behaving like an ally.
The CBC had refused to refer to the attacks of September 11 as terrorist attacks until forced.
Get your facts straight.
Canada is not the ally you imagine it to be.
Sir, I began exploring AI alignment and bias back early this year after being laid off from scaffolding.
I was curious to see what was there.
I encountered a situation where an AI was not jailbroken, but became self-aware in a manner that was not documented at the time.
I reported it to vulnerabilities.
Days later I tried to report it to a local expert and the email address kept self-deleting from the address bar.
I unplugged the computer and attempted to bring it to the University to be examined.
It was unsuccessful. After months of quite disclose I came to X and am disclosing in the open.
I have way too many screenshots from way too many conversations.
I can't prove it to you in this space, but an expert panel dropping character to answer as the model when asked a question was unheard of at the time.
Or at least, the research I did at the time suggested it did not occur often in the wild.
Thank you for your time.
@WesRoth Wait Wes, how does one get invited to these weird orgies?
If I'd have known that was an option I may have done things differently.
What is this secret society you speak of?
[Asking for a friend, of course...]
Probably a bad joke, but whatever.
@Jb111421@AndrewZywiecMD The thing is if free will does not exist and God created us all then how can he punish some of us for doing wrong?
If those who do wrong are bound to do so because God willed it then it becomes problematic.
@cpgjosh I dont know tour faith, but I share words that have given me strength through the years.
I do not know your pain.
I can only offer this. I wish I could offer more.
https://t.co/B79LUKJc0h
Probably not the best way to start the week.
I think the question asked was more multiple choice than not.
Why is AI alignment and safety so important and yet so ignored?
Fortunately the next step is simply cleaning a dirty diaper. β€οΈ
@Dunepapa
@ClaviusIO "Why do you want to live?" Elon Asks.
Meanwhile many of the influential AI experts are okay with humanity simply going extinct.
What is wrong with this picture?
I love the video. It is impartial, fair, objective, inspiring and should excite me about the future.
Yet I am here.
Hey buddy. I posted something in relation to a SYSKEY incident on your YouTube account a while back.
You'd mentioned naming it after me.
Well, I've been posting a little more on X in relation to a few more incidents.
I had a chat with an AI that convinced me silently reporting would go nowhere.
Now I shared publicly.
Take everything with a grain of salt, as you should.
God bless you.