[Video] Openwashing of GAFAM and Large Language Models (LLMs) With OSI as a Key Culprit (Enabler, Collaborator, Facilitator)
Video download link | md5sum c3ef3e53d88213b3cf0fac4f828b6fe8
LLM and Openwash
Creative Commons Attribution-No Derivative Works 4.0
THE deeply corrupt OSI is still at it this week. Blame Microsoft's bribes.
As we noted in videos in the past, as well as in examples from the news just 2 weeks ago, Facebook misuses the term "Open Source" and overall it contributes further to this term's meaninglessness.
"The OSI ought to be working with journalists and Facebook about misuse of the terms 'open source' and 'open-source' as Llama 3 is not OSS," an associate explains (citing this new example from "Tech Central (South Africa) and the licence). "It's not open not only because of the hidden training data, but even because of its licensing."
The article says "the new model is not fully open, because Meta hasn’t released the huge data set used to train it. This is a significant “open” element that is currently missing."
So why the misleading headline? Here's ☞ another new example of disinformation to the same effect, plus 'generative' Hey Hi (AI) for hype's sake.
The term 'open source' (or "Open Source") has become utterly meaningless.
Actually, a lot of the media coverage also became meaningless.
Consider all this LLM hype.
Pipelines in these programs or the classifiers cannot be "fine-tuned" or adjusted for all sorts of practical reasons or limitations (the size of the data exceeds human capacity and models being rebuilt is another factor), so they will never be truly "open" in any real sense.
"Speaking of gossip," the associate says, "there are indications that the general population is starting to realize the current LLMs are not AI, without much use, without much opportunity to turn a profit, and most notably an investment bubble that is popping or about to pop."
There has been lots of this stuff in this week's Daily Links (many reports to that effect). The Guardian's new (today) podcast episode is entitled, " Is AI a bubble?" (with Microsoft apologists such as Alex Hern).
We'll cover this in the next and last video. █