Live data from Hacker News

Jeff Dean interview: Machine learning trends in 2020

venturebeat.com

31–40 of 60 posts

Re: Jeff Dean interview: Machine learning trends in 2020

#31

I don't like the idea of a computer that can think for itself, I don't like the idea of computers will replace humans jobs, I don't like the way we are heading.

You have nothing to worry about :-) I would recommend that you read the book by Gary Marcus and Ernie Davis for a country viewpoint.

Re: Jeff Dean interview: Machine learning trends in 2020

#32
post #4

Somebody tell Jeff about Jevon's Paradox, or actually don't bother.

Is Jevons paradox really a problem when your carbon footprint is zero? > VentureBeat: One of the things that’s come up a lot lately, you know, in the question of climate change — I was talking with Intel AI general manager Naveen Rao recently and he mentioned this idea [that] compute-per-watt should become a standard benchmark, for example, and some of the organizers here are talking about the notion of people being…

even if the carbon footprint is zero (which it isnt) , all this energy is basically a large scale heater.

Re: Jeff Dean interview: Machine learning trends in 2020

#33
post #4

Somebody tell Jeff about Jevon's Paradox, or actually don't bother.

Jevon's Paradox [1] An example from [1]: ``` Jevons observed that England's consumption of coal soared after James Watt introduced the Watt steam engine, which greatly improved the efficiency of the coal-fired steam engine from Thomas Newcomen's earlier design. Watt's innovations made coal a more cost-effective power source, leading to the increased use of the steam engine in a wide range of industries. This in turn…

Jevon's paradox only occurs under very specific circumstances. I would bet on it not happening in the context Jeff Dean is referring to.

https://en.wikipedia.org/wiki/Jevons_paradox#Cause

> The size of the direct rebound effect is dependent on the price elasticity of demand for the good.[11] In a perfectly competitive market where fuel is the sole input used, if the price of fuel remains constant but efficiency is doubled, the effective price of travel would be halved (twice as much travel can be purchased). If in response, the amount of travel purchased more than doubles (i.e. demand is price elastic), then fuel consumption would increase, and the Jevons paradox would occur. If demand is price inelastic, the amount of travel purchased would less than double, and fuel consumption would decrease. However, goods and services generally use more than one type of input (e.g. fuel, labour, machinery), and other factors besides input cost may also affect price. These factors tend to reduce the rebound effect, making the Jevons paradox less likely to occur.

Re: Jeff Dean interview: Machine learning trends in 2020

#34

Earlier quoted context omitted.

It's a mix of (1) great work around sustainable processes and (2) nuances in defining carbon footprint. (disclaimer: work at G) https://storage.googleapis.com/gweb-sustainability.appspot.c... https://www.google.com/about/datacenters/renewable/

Buying renewable energy does not make your footprint zero. You can twist the definition until it suits your needs but if you properly count footprint of the chain, it cannot be zero.

Google buys high quality carbon offsets, not just renewable energy. It's not clear whether that makes the footprint zero. It surely depends on how you count. But their approach is less naive than "just buy RECs."

Re: Jeff Dean interview: Machine learning trends in 2020

#36
post #4

Somebody tell Jeff about Jevon's Paradox, or actually don't bother.

I think google as a whole understands this concept. Isn't that why they've held off on the release of their self-driving car? I can't remember where I read this from but I remember an interview where someone said they didn't want to release their self-driving car to the world until it was X-times better than the average human driver, and extremely competitive with its price.

Re: Jeff Dean interview: Machine learning trends in 2020

#37
post #26

This the first nonsense talk by Jeff Dean. AI doesn't help at all it battling climate change, only politics do help. The models are accurate enough for centuries, AI would only help in hard forecasting in the usual 2 weeks window on local events. Long term on global scale there's no AI needed at all. So it looks like he ran out of topics to entertain himself. Or he went politician. Which would be a welcoming change.

I’d agree that Jeff there is a man with only a hammer walking around telling us that Climate Change is a nail. Engineers can’t fix everything.

Of course the engineering is important, but politics is 1000% the road block on climate action and that is where efforts should be focused,

Re: Jeff Dean interview: Machine learning trends in 2020

#38

I don't like the idea of a computer that can think for itself, I don't like the idea of computers will replace humans jobs, I don't like the way we are heading.

I wouldn't like the way it was heading either if Netflix could manage to recommend a movie that I might actually like. As is, I'm not that worried.

Re: Jeff Dean interview: Machine learning trends in 2020

#39
post #5

Does anyone have tips on how a European based developer with machine learning expertise can get involved with projects battling climate change like Jeff is talking about here?

My current client is a sophisticated AI/ML startup/consultancy, Faculty ( https://faculty.ai ). They have extensive experience in a variety of areas, and do some cutting-edge stuff. If you're intested, either ping me (email in the profile) and I can connect, or use the website.

+1 for Faculty - very smart team

Re: Jeff Dean interview: Machine learning trends in 2020

#40
post #5

Does anyone have tips on how a European based developer with machine learning expertise can get involved with projects battling climate change like Jeff is talking about here?

We're doing some interesting work in this field at Cervest (based in London). https://www.cervest.earth/
Post reply on HN