- LLM wrappers are crumbling- these businesses will not survive as foundational models begin to offer the functionalities consumers have been seeking. This maybe represents a failure in interpretation of individuality as horizontal growth (these startups didn't last too long...), which brings up the next point.
- Individuality and ego are human needs. In the past year, there's been a boom in using AI to personalise products to consumers. A take on this: are companies really doing "individuality as horizontal growth" as consumers would like them to? Do consumers really need businesses to tell them that they're individuals? Are personalised feeds actually honouring their individuality, or are they just grouping consumers into demographics that homogenise them and strip away their individuality? And do businesses realise that consumers aren't blind to this? (reference at the rise & mission of Bluesky).
- Above connects to backlash from consumers in response to AI slop. AI slop is a feature of this era of conformity-> so look ahead to the era of individuality (reference historical back-and-forths between these two concepts- ie 1910s were the tail of social conformity and featured great social unrest until it was followed by the 1920s)
Also:
- I am assuming Hacker News is visited by a lot of software devs: one-shot learning and zero-shot learning could be something to check out! (reference Meta currently building Large Concept Models)
That's all. I hope you got value out of this post and have a good day!
(long note if you appreciated this and want to see an example of the above concepts and judge some ideas for yourself:
The above was written by me, Audrey Fei, I'm 16:
- I've been looking at the speech area of AI (text and image has already had its moment). currently have a patent pending for a real-time accent conversion technology (rooted in the idea of a “Golden Speaker” which dates back to 1990; also the patent is for the specific architecture of the System, which is different from standard architectures expected in the industry).
it's at loxai.tech (I'm trying to raise awareness to this tech so there's a waitlist page). the central idea behind it is that: at least in the subfield of speech processing, Golden Speakers are contrarian to the norm of speech synthesis (where speaker identities are lost, instead of preserved). also it uses one-shot learning (a MAML) in its structure
- there's a MVP that led to the above (neutboom.com). It centers around natural language acquisition (differentiates itself from the rest of the language learning market). It's might be a bit clunky in places (you can just tell me about anything that's ugly) but hopefully it's interesting or adds some perspective (at least personally I feel that translation is a middleman.....and why would you need a middleman especially when it overcomplicates the process of language learning? input on this would be appreciated).
)
Anyway I hope you have a creative rest of your day!