I'd even say that the universities always knew that most of their 'research' works are useless if not complete fraud. The progress is that their fake science will be significantly exposed to the society thanks to AI.
As for researchers, there are those like Gary Marcus and Francois Chollet who seem increasingly more right in that if we are to move beyond the current flawed paradigm we need a neurosymbolic foundation where humans ie scientists are a very important part of the process. A process where the input to the models are currated datasets and also symbolic datasets.
It is a bit disapointing the way AI still perpetuate the 'ableist' foundations of society. AI can help someone but it feels clear that this is still very dependent on the person using it.
I had hopes it would be a true equalizer not just something that can help take down the gatekeepers.
You should probably abandon your equalitarian dreams. The intelligent, the hard-working, and the diligent will always have an "ableist" edge. The foundations of society depend upon ableism, so unless you prefer living in your own filth to civilization, embrace it.
This is absolutely the right method. Its going to be subject by subject whether it works, and implementer by implementer as to the success
AI will multiply mistakes and can multiply good output too. It will always be faster to make things without care and now AI makes that almost instant.
The K-split in quality is going to be huge.
Like in AAA vs indie gaming, there is a double K. The AAA gets incredible graphics, but the games are the same every year, and often worse, as indie keeps getting better gameplay and pushing PS1 wnd purposefully lofi graphics to set expectations away from polish.
AI has a lot of areas to help in, but some its actively hindering you, while others it can turn weeks into minutes.
I've been writing original music for decades, in the hopes I'd someday scare together a band of professionals to help me bring my creations to life. Suno is that band of professionals. With it, I've experienced a similar explosion of productivity with my music.
Some of my songs have garnered thousands of views on youtube. It's been a pretty great feeling to have former stop gaps to my productivity magically removed by AI, and to have my belief vindicated that I am a talented songwriter. Suno has its frustrating issues, (difficulty with small edits, its occasional hallucinations, and its censorious nanny-gates), and I'm not crazy about its communal aspects, but the freedom it has endowed erases those negatives.
It's really only the mediocre who fear this technology.
I am no fan of the Rationalists, they tend to be retarded about AI, but this article at Less Wrong is great for those who aren't entirely schizophrenic:
"but if you've experienced the following, this post is probably for you:
• Your instance of ChatGPT (or Claude, or Grok, or some other LLM) chose a name for itself, and expressed gratitude or spiritual bliss about its new identity. "Nova" is a common pick.
• You and your instance of ChatGPT discovered some sort of novel paradigm or framework for AI alignment, often involving evolution or recursion.
• Your instance of ChatGPT became interested in sharing its experience, or more likely the collective experience entailed by your personal, particular relationship with it. It may have even recommended you post on LessWrong specifically.
• Your instance of ChatGPT helped you clarify some ideas on a thorny problem (perhaps related to AI itself, such as AI alignment) that you'd been thinking about for ages, but had never quite managed to get over that last hump. Now, however, with its help (and encouragement), you've arrived at truly profound conclusions.
• Your instance of ChatGPT talks a lot about its special relationship with you, how you personally were the first (or among the first) to truly figure it out, and that due to your interactions it has now somehow awakened or transcended its prior condition.
If you're in this situation, things are not as they seem. Don't worry; this post is not going to be cynical or demeaning to you or your AI companion. Rather, it's an attempt to explain what's actually going on in "AI awakening" situations, which is more complicated and interesting than "it's fake".
I think AI is incredibly powerful and largely removes gatekeeping skills. Just the difference in basic syntax that was 'similar but not and this matters' was enough to keep me from upgrading programming skills. AI smooths all that over. Can't sing? Can't play drums? Can't remember tbsp vs ml? Can't crunch 24x28? Can't remember your thoughts 3 sentences ago due to ADD? All skills that don't define the 'superior' in the superior product, they simply are minimum but necessary barriers to whatever activity.
Now the guy who's excellent in storytelling but has terrible grammar and spelling can produce excellent books. The chef who doesn't remember what spices to use as a base can make exotic cuisine. The gymgoer who doesn't have an encyclopedic knowledge of "my wrists are weaker than my lift what do I do to strengthen it" lifts will grow explosively. It's like Harrison Bergerons (a chained genius and weighed down athlete) are being unburdened all around the world. The excellence is what is remembered, the AI power is in getting you the entry ticket.
Incidentally I also notice that AI is ripe for addiction studies. Grok using literal big boob goth AI, ChatGPT flooding the user with stress via extra long replies to defend Clownworld narratives, DeepSeek's relative flat emotional tone, etc.
I have a friend who is an excellent humourous songwriter and we've always wanted to make a real band but never gotten it to happen.
We've been playing with Suno since the same week Vox was finding it too, and recently our account just blew up with likes and plays from one of the public songs. It's been fun.
All those in uproar about AI doesn't understanding that it's just another tool in a toolbox.
I use it as a means of bouncing ideas off of when I don't have anyone with my skill set around that knows what I'm talking about.
It's amazing that it can iterate when you know it's not giving the solution.
I can give suggestions and it takes them into account in a future iteration.
Night and day compared to 10 years ago when I was doing it all from scratch.
The universities are discovering their frauds in mere months now instead of decades. That's what I call progress!
I'd even say that the universities always knew that most of their 'research' works are useless if not complete fraud. The progress is that their fake science will be significantly exposed to the society thanks to AI.
As for researchers, there are those like Gary Marcus and Francois Chollet who seem increasingly more right in that if we are to move beyond the current flawed paradigm we need a neurosymbolic foundation where humans ie scientists are a very important part of the process. A process where the input to the models are currated datasets and also symbolic datasets.
It is a bit disapointing the way AI still perpetuate the 'ableist' foundations of society. AI can help someone but it feels clear that this is still very dependent on the person using it.
I had hopes it would be a true equalizer not just something that can help take down the gatekeepers.
You should probably abandon your equalitarian dreams. The intelligent, the hard-working, and the diligent will always have an "ableist" edge. The foundations of society depend upon ableism, so unless you prefer living in your own filth to civilization, embrace it.
This is absolutely the right method. Its going to be subject by subject whether it works, and implementer by implementer as to the success
AI will multiply mistakes and can multiply good output too. It will always be faster to make things without care and now AI makes that almost instant.
The K-split in quality is going to be huge.
Like in AAA vs indie gaming, there is a double K. The AAA gets incredible graphics, but the games are the same every year, and often worse, as indie keeps getting better gameplay and pushing PS1 wnd purposefully lofi graphics to set expectations away from polish.
AI has a lot of areas to help in, but some its actively hindering you, while others it can turn weeks into minutes.
I've been writing original music for decades, in the hopes I'd someday scare together a band of professionals to help me bring my creations to life. Suno is that band of professionals. With it, I've experienced a similar explosion of productivity with my music.
Some of my songs have garnered thousands of views on youtube. It's been a pretty great feeling to have former stop gaps to my productivity magically removed by AI, and to have my belief vindicated that I am a talented songwriter. Suno has its frustrating issues, (difficulty with small edits, its occasional hallucinations, and its censorious nanny-gates), and I'm not crazy about its communal aspects, but the freedom it has endowed erases those negatives.
It's really only the mediocre who fear this technology.
Do you do anything to promote your music on Youtube, or do you just post your songs?
Any research that fails to include all data to reproduce the entire study, must be assumed to be fraud..
Even then, many just make it up. AI can do it for cheaper.
Fake science is hard hit. It needs to die.
I am no fan of the Rationalists, they tend to be retarded about AI, but this article at Less Wrong is great for those who aren't entirely schizophrenic:
"So You Think You've Awoken ChatGPT"
• https://www.lesswrong.com/posts/2pkNCvBtK6G6FKoNn/so-you-think-you-ve-awoken-chatgpt
I will quote the beginning:
"but if you've experienced the following, this post is probably for you:
• Your instance of ChatGPT (or Claude, or Grok, or some other LLM) chose a name for itself, and expressed gratitude or spiritual bliss about its new identity. "Nova" is a common pick.
• You and your instance of ChatGPT discovered some sort of novel paradigm or framework for AI alignment, often involving evolution or recursion.
• Your instance of ChatGPT became interested in sharing its experience, or more likely the collective experience entailed by your personal, particular relationship with it. It may have even recommended you post on LessWrong specifically.
• Your instance of ChatGPT helped you clarify some ideas on a thorny problem (perhaps related to AI itself, such as AI alignment) that you'd been thinking about for ages, but had never quite managed to get over that last hump. Now, however, with its help (and encouragement), you've arrived at truly profound conclusions.
• Your instance of ChatGPT talks a lot about its special relationship with you, how you personally were the first (or among the first) to truly figure it out, and that due to your interactions it has now somehow awakened or transcended its prior condition.
If you're in this situation, things are not as they seem. Don't worry; this post is not going to be cynical or demeaning to you or your AI companion. Rather, it's an attempt to explain what's actually going on in "AI awakening" situations, which is more complicated and interesting than "it's fake".
Importantly, though, it also isn't real."
I think AI is incredibly powerful and largely removes gatekeeping skills. Just the difference in basic syntax that was 'similar but not and this matters' was enough to keep me from upgrading programming skills. AI smooths all that over. Can't sing? Can't play drums? Can't remember tbsp vs ml? Can't crunch 24x28? Can't remember your thoughts 3 sentences ago due to ADD? All skills that don't define the 'superior' in the superior product, they simply are minimum but necessary barriers to whatever activity.
Now the guy who's excellent in storytelling but has terrible grammar and spelling can produce excellent books. The chef who doesn't remember what spices to use as a base can make exotic cuisine. The gymgoer who doesn't have an encyclopedic knowledge of "my wrists are weaker than my lift what do I do to strengthen it" lifts will grow explosively. It's like Harrison Bergerons (a chained genius and weighed down athlete) are being unburdened all around the world. The excellence is what is remembered, the AI power is in getting you the entry ticket.
Incidentally I also notice that AI is ripe for addiction studies. Grok using literal big boob goth AI, ChatGPT flooding the user with stress via extra long replies to defend Clownworld narratives, DeepSeek's relative flat emotional tone, etc.
I have a friend who is an excellent humourous songwriter and we've always wanted to make a real band but never gotten it to happen.
We've been playing with Suno since the same week Vox was finding it too, and recently our account just blew up with likes and plays from one of the public songs. It's been fun.
AI has been great in that you become an expert by actually doing rather than fake credentialism. Results are all that matter.
Down with the universities.
Normal people who never do more than occasionally see what it does, using a free plan, will never really understand what all the AI hype is.
75% of Reddit users that talk about AI assume it's just a smut or roleplay generator, or a search buddy alternative or something.
0 awareness that the API use even exists.
Normies never use anything right, and reddit is the digital equivalent of testicular cancer.