RE: https://mstdn.ca/@teledyn/116652708401285794
-
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
@timnitGebru walks among us.
-
-
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit@harold if anyone just sits in on a lecture in first year computer science... none of this is a new idea, it's only being implemented now at scale. The hardware has made it feasible. It's a race to the plateau.
You can't train it on it's own junk. Everything about the situation is very well known, and No, it's not all a fresh discovery of Today.
That's just the HYPE-FACTORY version saying 'ITS A NEW WORLD' etc.
-
@harold if anyone just sits in on a lecture in first year computer science... none of this is a new idea, it's only being implemented now at scale. The hardware has made it feasible. It's a race to the plateau.
You can't train it on it's own junk. Everything about the situation is very well known, and No, it's not all a fresh discovery of Today.
That's just the HYPE-FACTORY version saying 'ITS A NEW WORLD' etc.
-
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
RE: https://mstdn.ca/@teledyn/116652708401285794
"every single warning that paper made about large language models has now happened at scale"
1. The hallucination problem before anyone had a word for it.
2. Bias amplification
3. Environmental cost
4. Documentation — the training datasets being assembled were too large for anyone to actually audit -
M monkee@other.li shared this topic