@AndrewG29922648@AlchemyAmerican I kept rereading the paper looking for some novel spark because of exactly this, the obviousness of it all.
My conclusion is this is his personal philosophy dressed up with a LOT of big words so as to sound more convincing. The math has very little to do with his premise.
we are being gaslit about AI on a societal level. Everybody is vibe coding but I haven’t seen one useful thing get produced. Everybody has agents doing something but nothing useful is getting done. Cool you had AI summarize a PDF and make a template. Nice
Their first level of answers are “internet advice” which is why the newer models are so important… you can call that out and have them look into the reality of something instead.
It’s like a hopeful actor googling “how to succeed as an actor in Hollywood”… the internet advice answer is “practice and go to lots of auditions” but the reality is much more complex.
If someone only sees and follows the internet advice answer, they will fail and wonder what’s wrong with them.
But in reality they didn’t even fully understand the the game they were playing or how to even try and win.
Respect for the time spent developing his creative ideas, but they are just that, his own creative ideas.
His model doesn’t actually prove anything or reveal any new information. His math does not show something that wasn’t understood.
It is his own description of what already exists, with a lot of his own random ideas inserted in.
This guy is doing the equivalent of putting humans under AI psychosis. He speaks confidently about things that seem hidden and people buy into it.
@venturetwins@plotpartyai@watchmochi@techreview Right now there is a ton of one and done characters and plots. I think recurring characters / world and potentially valuable IP is a bigger play here
@ihtesham2005 A counter to this is… how often where the student writing by hand in their daily lives? Probably very rarely. So writing WOULD activate parts of their brain that wouldn’t be activated if they wrote all day every day
I really, really wish that were true! I wish I could somehow brute force my way into simply overcoming my personal shortcomings with the power of love!
It is the same thing so many kids hear everyday, “if you just paid attention” “if you cared about your grades you would have remembered your homework!” “You can pay attention to the thing you’re interested in so I know you CAN pay attention and remember you choose not to!”
But for someone with executive function issues it is quite literally like telling someone missing a leg that they just don’t really want to walk enough.
This is an executive function prosthetic, a first look at a new kind of help for people who really need it.
I don’t expect everyone to understand how powerful and needed that is!
Just bringing the vibe back to two people talking, X makes people default to yelling at perceived enemies.
It does sound like the fine / not fine line is already pretty vague even in this discussion and will continue to get more vague.
Ie, I’m talking to a recorder, fine, the “AI” reminds me what I said later… grey or possibly evil.
Navigating a crowd’s ever shifting morality is a losing battle I don’t try to attempt
@Rebel_Fraggle@CharlotteAlter This seems important to you. What separates writing down “my wife likes Shoe X,” from saying “my wife likes Shoe X, remember that” to you?
In my case it is referring to a system I built for myself. It uses hardware devices I built that record and stream audio to a home server which transcribes everything. I work alone from my home so I speak out loud throughout the day.
Those transcripts are scanned for any commitments I’ve made, things I need to do, stuff with a deadline, etc. this is why I am talking out loud, so I can get it out of my head and onto “paper”.
They are also searched for information from different life domains. One for each family member, one for my business, one for my family as a whole, one for my main hobby…
And each of those domains have short and long term goals set by me. Each night, this information is compiled in each domain along with an outlook of where I am at with these goals. This also includes “working time” for each domain to do what it can to help move me towards my goals.
For example, it noticed my son’s event was cancelled one evening and suggested I take my daughter to her dance class that night and take her out for ice cream afterwards, because we hadn’t spent one on one time in a while.
It is only able to do this because I designed to do this. I point that out because, to those not deeply versed in building with LLM’s it can seem like they can just do things on their own. There is nothing like this prebuilt for other people, to my knowledge.
I had planned to release it but have instead changed to simply using it myself and pouring into my current work.