News Expert pours cold water on Apple's downbeat AI outlook — says lack of high-powered hardware could be to blame

[Admin]

Contemporary AI models are not silly, Apple just does not have proper hardware to test their limitations, says professor Seokjun Kwon of Sungkyunkwan University.

Expert pours cold water on Apple's downbeat AI outlook — says lack of high-powered hardware could be to blame : Read more

 

[coolitic]

"Experts" that have a vested interest in non-stop AI-interest-churn.

Andrew Kelley had a good take on this:

LLMs are a way to make software take orders of magnitude more computational power, electricity, and human labor, while delivering a product whose extremely volatile quality is impossible to assure. The work will never be completed; it will only create the need for ever more labor.

From: https://andrewkelley.me/post/why-we-cant-have-nice-software.html

 

[Penzi]

The fundamental nature of token-based LLMs strikes me as having an in-built limitation that likely conforms to Apple’s take rather than NEED MORE POWER

 

[bigdragon]

I don't think this professor is being objective towards Apple's conclusions. I work in the "real world" and in an industry the AI tech bros are trying to stuff all sorts of AI crap into. LLMs are not trustworthy. I've seen Apple's research confirmed too many times to count. The inclination of these AI solutions to hallucinate and make up facts to support their conclusions makes them objectively dangerous in some scenarios. I'm honestly shocked a few companies haven't gone under yet for relying too much on AI.

This professor should work on fixing the hallucination problem. LLMs hallucinate no matter how much hardware and performance you throw at them. Begging Apple for more hardware performance is like begging for 1 more lane on a congested highway or 1 more turn after losing a game of Civilization -- hardware performance only masks a fundamental flaw in the system.

 

[Lieutenant Barclay]

Shill pours cold water on Apple's downbeat AI outlook​

Fixed it.