Google Wizards: Hey, We Knew But Did Not Intervene. Very Bard Like
February 15, 2023
I read two stories. Each offers a glimpse into what I call backing away and distancing. I think each reveals the failure of Google governance. You may disagree. That’s okay, particularly if the stories are horse feathers. My hunch is that there is a genetically warped turkey under the plumage.
The first item is from the increasingly sensational Insider. The story is “Google Didn’t Think Its Bard AI Was Really Ready for a Product Yet, Says Alphabet Chairman, Days after Its Stock Fell Following the Chatbot’s Very Public Mistake.” The write up pivots on information (allegedly 100 percent dead solid in the bull’s eye) provided by John Hennessy, the chairman of Alphabet. The chair person! What did this captain of the digital titan say? I quote from the write up:
“I think Google was hesitant to productize this because it didn’t think it was really ready for a product yet, but, I think, as a demonstration vehicle, it’s a great piece of technology….He added Google was slow to introduce Bard because it was still giving wrong answers.
From my point of view, isn’t the role of the Board of Directors, and specifically the Chair, supposed to provide what might be called governance guidance? Since this admission of “giving wrong answers” is made public after the disaster in a city where a great lunch is easy to obtain, I would suggest that the bowl of soupe a l’oignon was prepared from a bag of instant convenient food: Not particularly good but perfect for a high school science club snack.
The second item is from CNet, which has some experience with smart software. The article is “Computing Guru Criticizes ChatGPT AI Tech for Making Things Up.” And who is the computing guru? None other than Vint Cerf, one of the father’s of the Internet if I remember something I heard at a conference.
The CNet article reported as actual factual:
But, speaking Monday [February 13, 2023] at Celesta Capital’s TechSurge Summit, he did warn about ethical issues of a technology that can generate plausible sounding but incorrect information even when trained on a foundation of factual material. If an executive tried to get him to apply ChatGPT to some business problem, his response would be to call it snake oil, referring to bogus medicines that quacks sold in the 1800s, he said. Another ChatGPT metaphor involved kitchen appliances.
Then this allegedly accurate quotation from the father of the Internet and Google guru:
“It’s like a salad shooter — you know how the lettuce goes all over everywhere,” Cerf said. “The facts are all over everywhere, and it mixes them together because it doesn’t know any better.”
Did the Googlers crafting Bard run the demonstration by Mr. Cerf? Nope. The write up says:
Cerf said he was surprised to learn that ChatGPT could fabricate bogus information from a factual foundation. “I asked it, ‘Write me a biography of Vint Cerf.’ It got a bunch of things wrong,” Cerf said. That’s when he learned the technology’s inner workings — that it uses statistical patterns spotted from huge amounts of training data to construct its response. “It knows how to string a sentence together that’s grammatically likely to be correct,” but it has no true knowledge of what it’s saying, Cerf said. “We are a long way away from the self-awareness we want.”
It seems to me that if the father of the Internet is on staff, it would make sense to get some inputs.
Let’s recap:
- After the fact, the Chair of the Board points out known problems but does not invoke action based on the need for governance related to product performance. Seems like something slipped betwixt the cup and the lip.l
- After the fact, the father of the Internet points out that he was “surprised” that Google technology generated misinformation. Again … after the fact.
Is the company managed by responsible adults or individuals who believe themselves to be in a high school science club? Are Googlers indifferent to the need to get their act together before they take the show on the road.
I think the French could label either Googlers’ comment as observations offered in l’esprit de l’escalier. Accurate but not management.
Stephen E Arnold, February 15, 2023
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
