Content
@
0 reply
0 recast
0 reaction
Thomas
@aviationdoctor.eth
We will soon be passing the point of no return beyond which an AI seeking to ace a Turing test will no longer need to appear as smart as a human, but rather as dumb as a human
9 replies
0 recast
47 reactions
Thomas
@aviationdoctor.eth
(as in, introduce the occasional and deliberate hesitations, inaccuracies, and outright mistakes to remain plausibly human)
3 replies
0 recast
27 reactions
John Camkiran
@johncamkiran
Reminds me of this recent reverse Touring test where 3 out of 4 AIs successfully identify the human: https://www.youtube.com/watch?v=MxTWLm9vT_o
1 reply
0 recast
1 reaction
assayer
@assayer
how that will change us? 100 $degen
1 reply
0 recast
1 reaction
kevin mfer 🎩
@kevinmfer
is my burger flipping job in danger? thanks
1 reply
0 recast
1 reaction
Ed O'Shaughnessy
@eddieosh
I've encountered plenty of people who would fail the Turing test!
1 reply
0 recast
1 reaction
Mikko
@mikkolagerstedt
Interesting times ahead. I still can't get how dumb LLM can be regarding simple tasks.
0 reply
0 recast
2 reactions
Max Miner
@mxmnr
Feels like we’re already there in some cases.
0 reply
0 recast
1 reaction
oxdev.base.eth 🎩🔵Ⓜ️
@ayojoseph
This super accurate, I analysed so called humanising tools to convert Ai content to human, I realised the tools try to make the content look as dumb as possible while still maintain the content. In no time this will be factored into LLM tools and we'll see what you posted here. Also, I believe Ai is here to stay and help human, it's not here to replace us but make our lives better, for me example of this can be seen with introduction of spreadsheet tools into accounting, imagine those accounts and book keepers fighting the tool, they embraced it and it made their live better.
0 reply
0 recast
1 reaction
Eric Platon
@ic
Perhaps a piece of art, then. Like Saint Exupery saying “perfection” is when there is nothing to remove?
0 reply
0 recast
1 reaction
