"Artificial intelligence researchers are calling out both CBS and Google for overhyping AI after the network aired a high-profile 60 Minutes interview with Google CEO Sundar Pichai on Sunday. In the segment, correspondent Scott Pelley claimed that an AI program made by Google had learned a language it had never seen before all by itself, and Pichai referred to AI tech as a “black box” that even the people who worked in the field didn’t fully understand." https://www.buzzfeednews.com/article/pranavdixit/google-60-minutes-ai-claims-challenged
"“We discovered that with very few amounts of prompting in Bengali, it can now translate all of Bengali,” James Manyika, a Google vice president also interviewed by 60 Minutes, said on the segment. “So now, all of a sudden, we have a research effort where we’re now trying to get to a thousand languages.”
This is how their hype undermines startups actually serving their communities, like @asme's lesan.ai.
@meg, a researcher and ethicist at AI startup Hugging Face who formerly co-led Google’s AI ethics team, pointed out that PaLM, is, in fact, trained on Bengali according to a paper published by Google’s own researchers. The paper says that Bengali made up 0.026% of PaLM’s training data.
"“By prompting a model trained on Bengali with Bengali, it will quite easily slide into what it knows of Bengali: This is how prompting works,” Mitchell tweeted. It is not possible, she added, for AI “to speak well-formed languages that you’ve never had access to.”"
"@emilymbender tweeted that the term “‘emergent properties’ seems to be the respectable way of saying AGI,” which stands for artificial general intelligence, a hypothetical technology that can learn on its own and perform tasks better than humans. “It’s still bullshit,” she said."
"“When tech leaders muddy the waters about how the technology actually works and invite us to believe it has mysterious ‘emergent’ properties, that makes it harder to create appropriate regulation,” Bender said."
@rysiek@timnitGebru Imagine if an executive in any other engineering field said that about their product.
"We don't know how it works but we're going to sell it to you anyway," is a pitch that literally only a Silicon Valley exec could deliver.
And also it's total bs. We know how these models work. They're big math equations whose weights are calculated from a training set to map inputs to outputs. It's not like it's magic or even quantum physics. The hype is ridiculous (but very profitable).
it might not reassure you to learn that this is how a lot of medical research is done. Metformin was being prescribed to diabetics and prediabetics for a decade or more before researchers found out how it works. (it alters the behavior of gut bacteria.)
@rayckeith there's a world of difference between "we don't know how this works, but this thing has been shown in medical studies to help with your particular ailment" and "we don't know how this works but we're going to push that thing down everyone's throats whether they like it or not and hype it up so that we sell some stock".
Add comment