No, because they already have LaMDA which is the same thing, a language model for dialogs.

Google could implement that into Google today, but the reason they don't is Google is a 200b/year business. You don't want to mess around with a business of that size for the fun of it.

We will probably see it in Google Assistant before they add it too Google.

OpenAI doesn't have this worry so they can put their model out into the public and if anything goes bad they don't lose anything. Google does.

That was my thought process too, particularly the business part. I find it hard to believe Google would want to give that up so easily. Will be interesting to see how it unfolds.

I think that could explain exactly why it's a threat to Google (thus yes to deep waters). Google likely has the tech/ability, but they have an existing business to protect which is a strong disincentive to experiment. It creates a strong opportunity for the upstart trying to carve out new territory.

Wonder if Google is skunkworking something like this at arms' length as a backup plan?

Also some Google employees have mentioned on HN that Google has been analyzing similar things and the compute is currently too expensive to make it profitable.

That's the real reason. And the fact that it decreases ad views.

sounds very monetizable. any question that can be solved with a product.. replace that with an affiliate link. maybe only a 5th of the queries will be monetized, but the ones that are are so specific/targeted that the ROI/click through will be great

Innovator dilemma in full swing.

ChatGPT can still hallucinate stuff, so if you want to write anything of substance, you have to check its claims through Google.

But Google is probably about to be massively poisoned with ChatGPT output. And things like this repo itself may get fed back in to future versions of GPT possibly impacting it over time.

That's a trippy thought. We're essentially writing real-time into the future GPT's memory.

Google doesn’t do anything to validate the search results lead to accurate information

And yet that doesn't mean Google is regularly returning wrong answers like Chat got is.

Don't put out this intellectually lazy argument without considering the magnitude of error and user ability to mitigate it.

Give it some time and the Internet will be swamped by AI-generated dubious crap, and at that point Google will lose its advantage.

Oh, that I completely believe in - ChatGPT isn't going to be the saviour of writers, it's going to be used to flood the internet and social media with autogenerated propaganda and submarine marketing in a scale never seen before.

The „news“, especially in finance, is more and more autogenerated these days

If you have an opinion I wouldn't mind hearing it. As I say, I haven't come to a conclusion myself but maybe because I lack the information/knowledge about it.