> OpenAI got to iterate, but maybe Google instantly move 37'd conversational AI?
If you write such speculative theories, I want to do, too. :-)
What if Google already "37'd" OpenAI, but the result is too dangerous to be released to the public, i.e. it takes a huge amount of work to "defuse"/"cripple" their AI for it to be suitable for "public consumption".
What if Google has an actual GAI that is so good they’ve secretly turned over all strategic decisions to it, and it is directing this “appear to be behind” strategy for reasons known only to itself?
If anyone was going to create Samaritan [1] and end the world, of course it would be Google. The AI would probably bring about the apocalypse trying to sunset the main search engine in a misguided fit of madness brought on by analyzing Google's past behavior [2]
> However, DeepMind considers Sparrow a research-based, proof-of-concept model that is not ready to be deployed, said Geoffrey Irving, a safety researcher at DeepMind and lead author of the paper introducing Sparrow.
> “We have not deployed the system because we think that it has a lot of biases and flaws of other types,” said Irving. “I think the question is, how do you weigh the communication advantages — like communicating with humans — against the disadvantages?
Seems less speculative than a company with more resources, more know-how, more talent, more training data, more compute, more product experience, more anything, somehow unable to ship a competing product. But that seems the common conclusion people draw.
If you write such speculative theories, I want to do, too. :-)
What if Google already "37'd" OpenAI, but the result is too dangerous to be released to the public, i.e. it takes a huge amount of work to "defuse"/"cripple" their AI for it to be suitable for "public consumption".