Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What can a technologist do about climate change?

worrydream.com

Post reply on HN
Print view View original

201–210 of 293 posts

Re: What can a technologist do about climate change?

#201

Post by cies »

pjc50 wrote: post #191
cies wrote: post #111

Great piece. I was thinking: "what programmers can do?" And came up with the "use compiled languages" answer. I compare Ruby and Haskell, as that's what I have experience with; but I expect it to hold true for any Ruby/Python/Java/etc vs C/C++/Haskell/Go/Rust/etc shootout. Just some typical numbers: Disk space to house one app deployment: 800MB Ruby/Rails vs 80MB Haskell/Yesod. Memory consumption of one instance: 350…

800MB Ruby/Rails What, really? How much of that is the app source? As a byte-shaving embedded developer I have zero experience with Rails. On the other hand, unless you're scaling really big, the CO2 cost of the computers is probably dwarfed by the cost of the developers commuting to work.

> How much of that is the app source?

Little. Most is libs.

> [...] unless you're scaling really big, the CO2 cost of the computers is probably dwarfed by the cost of the developers commuting to work.

Yups. That's why I mentioned "scale". Personally I cycle 7 mins to work :)

Re: What can a technologist do about climate change?

#202

Post by hobolord »

mouly wrote: post #189

Bret Victor has provided a first principles approach to solving climate change. I work at SolarCity we take a similar systems thinking approach and tackle the problems at their fundamental level. We have been vertically integrating the business from manufacturing highest efficiency solar panels, mounting systems that enable fast installation, electrical systems that can work with the grid, financing for homeowners th…

I applied to SolarCity for similar reasons, because I'd like to hopefully do something about climate change. Hope I hear back soon

Re: What can a technologist do about climate change?

#204

Post by orangecat »

brandmeyer wrote: post #28

I have a controversial opinion on this topic: I don't think it is morally ethical to use inefficient languages in high duty-cycle applications. By "high duty-cycle applications" I specifically refer to environments which consistently consume the bulk of available CPU resources: big web servers, caching intermeddiate servers, etc. The common rationale is to trade off engineering time for CPU time. But that trade also…

You're going to need actual numbers for this to be a valid argument. Data centers use only 3 percent of global electricity production (http://www.datacenterknowledge.com/archives/2014/12/17/under...), and it seems unlikely that a noticeable portion of that is caused by CPU-intensive software that isn't already well-optimized.

Re: What can a technologist do about climate change?

#205

Post by danieltillett »

DennisP wrote: post #185
danieltillett wrote: post #155

Earlier quoted context omitted.

>I'd like to see specific numbers on how that difficulty translates into carbon emissions. So would I. Getting good data in this area is near impossible. For something complex you can use the rough rule of thumb that the dollar cost is proportional to the emissions - as complexity increases the actual emissions converge towards the economy wide average. On this basis the emissions from building a current generation n…

Nuclear plants cost a lot up front but they produce a lot of power for a long time. Per MWh they're not that expensive compared to renewables. Taking the government's estimates for capital cost, we have onshore wind at $58/MWh, nuclear at $70/MWh, hydro $71, advanced coal $77, solar PV $109, offshore wind $168, solar thermal $192. But I don't think anyone would claim that the construction of a modern coal plant has s…

How many recent Nuclear plants have been built that will be able to generate power at $70/MWh? All the recent ones have had massive cost and time overruns.

This report from the UK is interesting reading [1].

1. http://www.parliament.uk/documents/post/postpn_383-carbon-fo...

Re: What can a technologist do about climate change?

#206

Post by spenczar5 »

chaosphere2112 wrote: post #133

I work on climate software at Lawrence Livermore National Laboratory[1]. My team works with big data and visualization. We support ACME[2] (the Department of Energy's climate modeling program), PCMDI[3] (a team that's been evaluating climate models since 1989), and a number of other projects and groups. If you are interested in what you can do for the climate, we need all the help we can get. Our code may not be the…

How comfortable are government software groups like this with remote workers?

Re: What can a technologist do about climate change?

#207

Post by noname123 »

magicalist wrote: post #200
noname123 wrote: post #198

Earlier quoted context omitted.

> Re: Malaria, homelessness, HIV, refugees, war, climate change, income inequality, gender inequality, hate crimes. There are tons of informatics needs for modeling malaria genome evolution specially against emergence of front-line drug resistance that is spreading from SE Asia to Africa. Tools need to be built to inform public health officials where the genome's gene flow is headed next to modify the front-line trea…

that's just stating that there is a need, not where to start :) I wouldn't have any idea on where to go to help out with those either

1) Join a lab as a programmer (e.g., National Weather Service (climate), J. Craig Institute (infectious diseases))

2) Read the computational papers on a subject you are interested; replicate the results in the papers and open source your software/pipeline; apply the method to a newer data-set.

3) Contribute to a open source informatics toolchain used for the subject (e.g., https://github.com/bigdatagenomics/adam)

Re: What can a technologist do about climate change?

#208

Post by DennisP »

pjc50 wrote: post #119
guscost wrote: post #13

While I don't really agree with the premise, my initial impression is that this should get ten upvotes for being so well reasoned and executed. Amazing work, I'm confident a lot of good will come of it. One recommendation, at your earliest convenience do something about this: > I didn’t mention nuclear because I don’t know much about it.

Nuclear generates a lot of controversy. On paper it looks great; in practice there it turns out to be very expensive or have nasty problems buried in the details. For example, molten salt reactors have corrosion and reprocessing problems. I don't blame the author for refusing to take a position on nuclear.

The Oak Ridge molten salt reactor lasted four years without serious corrosion. Terrestrial Energy and Thorcon are working on systems with sealed reactor cans that get replaced in similar timeframes.

Reprocessing isn't a necessity; it extends fuel supplies but even Terrestrial Energy's simple burner design without reprocessing is six times as fuel-efficient as conventional reactors. Running on thorium is more complicated, but all the molten salt companies other than Flibe are starting with uranium.

Re: What can a technologist do about climate change?

#209

Post by tavert »

emcq wrote: post #182
nabla9 wrote: post #15

>Here’s an opinion you might not hear much — I feel that one effective approach to addressing climate change is contributing to the development of Julia. Julia is a modern technical language, intended to replace Matlab, ... Well that was awkward transition.

Completely agree. I like Bret's stuff and love Julia but that section seemed very weird and almost preachy. He seems a little out of his element because a lot of physicists doing earth modeling use compiled languages like Fortran and C/C++ for the big simulations and python for the smaller stuff. Julia is cool though and a great substitute for the python stuff. It won't implicitly solve distributed computing and the…

> It won't implicitly solve distributed computing

Why not? It's out to explicitly address distributed computing - http://docs.julialang.org/en/latest/manual/parallel-computin...

Julia is performance-competitive with C++ and Fortran, and productivity-competitive with Python. It's out to do both. Julia code compiles to LLVM and is incredibly "close to the metal" including SIMD, ref http://docs.julialang.org/en/latest/manual/performance-tips/... and https://software.intel.com/en-us/articles/vectorization-in-j..., and GPUs http://blog.maleadt.net/2015/01/15/julia-cuda/

Try using Julia a bit before you talk about what it "won't solve."

Re: What can a technologist do about climate change?

#210

Post by likeapub »

grenzbegriff wrote: post #115
Jemaclus wrote: post #57

I've asked this question a few times but haven't really gotten a satisfactory answer. There are a lot of what I might call "epic problems" out there. Malaria, homelessness, HIV, refugees, war, climate change, income inequality, gender inequality, hate crimes. I don't know anything about creating vaccines or cures for malaria or HIV. I don't have a degree in social work or behavior science. I don't have any experience…

I think about the same things and ask the same question. My current plan is to save most of my money so I can reach financial independence very early. I'm 25 now, I'll be financially independent by 27, meaning I'll have enough money invested to live indefinitely on interest/dividends. My total expenses/spending is $25k/year and I'm sure I can keep them there if I want to. Most people who start making a silicon valley…

how much savings will u have by 27? do you have a house too?
Post reply on HN