[danielscrubs]

I believe you, but just for the sake of it: What AI research has had business value from Google in the last years that you could sell?

[FartyMcFarter]

WaveNet is one example:

"Google Assistant adds 9 new AI-generated voices": https://venturebeat.com/2019/09/18/google-assistant-gains-9-...

Data center cooling is another one:

"How Google is Using AI for Data Center Cooling": https://www.bmc.com/blogs/data-center-cooling/

[Isinlor]

AlphaFold may bring millions if not billions to DeepMind, Google and Alphabet. Figuring out structure of a single protein may cost up to 100 000 dollars.

Bert as applied to search from late 2019: https://www.google.com/amp/s/blog.google/products/search/sea...

[Rochus]

Time will show; winning the competition was just the first step.

[Isinlor]

Indeed, that is why I'm saying "it may", but people are crazy if they expect that basic research will make billions of profits a year after.

It normally takes some 10 to 20 years for basic research to produce profits.

mRNA vaccines are technology from 1990 and are first time used commercially on mass scale on humans 30 years later.

[karmasimida]

BERT has been powering all Google's search traffic at this moment:

https://searchengineland.com/google-bert-used-on-almost-ever...

It is THE biggest change that Google's search algorithm had ever been through, I would assume. And to push such a fundamentally different model to ALL their English traffic is pretty telling itself that how much an improvement Google had been seeing.

This is easily billions of ROI for Google, if not tens of billions.

[tinyhouse]

Are you kidding me? Everything they do pretty much. Google search results, Gmail Compose, Google Translate, etc etc.

[sgt101]

Alphafold is transformational for life sciences. I find it hard to articulate how much it's worth - maybe the sum of the top three life sciences companies today?

Honestly - the discussion is over, the AI folks won.