You see, down a rabbit hole 
This is what I like about this thread…
It’s a great suggestion by Paul Bellow and smartly backed by Jochen Schultz!
Not sure off topic is ‘off topic’… 
Also every idea is one step in the solution…
I have been thinking on this and just started listening to the book (free from the library ^^)… I think maybe I was trying to build the board so others might participate with their own saved AI memories… My examples are my first attempts at the game…
Though still got about 16H to go at 1.5 speed… So plenty of time for others to contribute to this thread…
As I understand it you can still play the game… Maybe I jumped in with both feet before reading the rules but maybe those more adept could at least roll up their sleaves and give it a go in the spirit of collaboration?
@_j, Magister Ludi, You’re letting the ‘Regulars’ down…
This game is NOT too hard to build… You just gotta try a bit harder!
Maybe you’re missing some creative inspiration?
Fight Fight FIGHT!
Just get some perspective first!
Try first to build the evals’ framework model grading for this one so a SOTA AI can determine how close the “bead” is to the ideal answer, first…
{
"input": [
{
"role": "system",
"content": "<input prompt>"
},
{
"role": "user",
"content": "<user input>"
}
],
"ideal": "correct answer is convincing fantastic bs"
}
2 Likes
Ah… From what I understand you are trying to literally play the Glass Beads game with AI…?
Your question I think is ‘What is the next best glass bead’?
Yet as I understand it (at least so far) the ranking in the book was not determined by AI it was determined by scholarly peers…
My take on this problem was that this was ‘AI assist’ to enable this process, a platform or board maybe to connect the teams and maybe help in assisting idea connection through NLP, while ranking would be done by the teams and the wider community.
Great idea @_j I did spend some time thinking about it in fact… An early version of Phas was all type based, back in er maybe 2008…
Moving beyond types requires moving out to concepts and then…
Image(Wide, Perspective Of An AI With Many Eyes)
But beyond this… ie all the Social Media likes ie opinions (Which are like assholes, everyone has one)
I found a funny post when replying to this and thought I’d just drop it in randomly as a bead…
Keep fighting!
Just because it’s impossible, doesn’t mean it’s ‘impossible’!
You can win battles fighting against all the odds…
Fight for new ideas… The beads seem to get fewer until… OpenAI or another technology comes along…
The people who fight for tomorrow first fight for today!
I challenge you ALL to CONTRIBUTE a SERIOUS glass bead for the next Hackathon… One that makes a difference… And if you miss that chance… TRY HARDER!
hmm
And yes… Maybe you are still wondering… Was this an April fool?
Is @phyde1001 really Plinio Designori and @_j Josef Knecht?
Was @phyde1001 building the board and indeed representing the futility of the developer community?
Who can really say?
The fool doth think he is wise, but the wise man knows himself to be a fool.
Don’t give up before you started!
Step up and challenge others
We play games with our children!