• 0 Posts
  • 64 Comments
Joined 2 years ago
cake
Cake day: July 1st, 2023

help-circle


  • Most large corporations’ tech leaders don’t actually have any idea how tech works. They are being told that if they don’t have an AI plan their company will be obsoleted by their competitors that do; often by AI “experts” that also don’t have the slightest understanding of how LLMs actually work. And without that understanding companies are rushing to use AI to solve problems that AI can’t solve.

    AI is not smart, it’s not magic, it can’t “think”, it can’t “reason” (despite what Open AI marketing claims) it’s just math that measures how well something fits the pattern of the examples it was trained on. Generative AIs like ChatGPT work by simply considering every possible word that could come next and ranking them by which one best matches the pattern.

    If the input doesn’t resemble a pattern it was trained on, the best ranked response might be complete nonsense. ChatGPT was trained on enough examples that for anything you ask it there was probably something similar in its training dataset so it seems smarter than it is, but at the end of the day, it’s still just pattern matching.

    If a company’s AI strategy is based on the assumption that AI can do what its marketing claims. We’re going to keep seeing these kinds of humorous failures.

    AI (for now at least) can’t replace a human in any role that requires any degree of cognitive thinking skills… Of course we might be surprised at how few jobs actually require cognitive thinking skills. Given the current AI hypewagon, apparently CTO is one of those jobs that doesn’t require cognitive thinking skills.


  • I don’t know, but it’s bad.

    At this point even our best case scenarios are still pretty bad; barring some massive breakthrough in carbon sequestration tech.

    And the “business as usual” scenarios are down right scary, millions of deaths annually. Never mind the economic consequences.

    In my other comment I talked about what needs to happen on the macro level.

    But the micro level is another story.

    I’m worried because the paths to mitigating the worst of it depend mostly on countries, people, corporations etc… making major changes to drive reductions.

    I seen the strategies the big companies have… they’re not coming close to making the difference needed. And the small companies aren’t even trying to measure their emissions let alone reducing them. It’s that lack of data that’s a part the problem. The data needed for decisions at the micro level isn’t available. It’s difficult to even identify what changes to make because you don’t know what impact a change might have outside of your control.

    So far it means we haven’t even got emissions to start going down. At best, they’ve just slowed the rate at which they’re going up.

    Governments should be pushing harder to mandate emissions reporting, but it’s politically unpopular so we’re still largely guessing about what decisions to make and that’s what leads to us all pulling in different directions making little progress.


  • (Sorry for the length here… this is actually my shortened version)

    89% of climate change is because we took carbon that was permanently sequestered underground in the form of oil, gas, and coal and burned it for cheap energy. We need to stop that entirely but you can’t “just stop oil”, you need to remove the demand not try to disrupt the supply.

    There are 4 broad strokes to making that happen:

    1. We need a metric fuck ton more carbon-free electricity generation asap. Not just enough to replace all existing fossil fuel-based electricity generation, but enough to supply double to triple the current generation capacity. Only about a quarter of the energy we get from fossil fuels is used to generate electricity, so as we switch things over to electricity, demand will increase exponentially.

    Renewables are great and we need to build as much as we possibly can, but what people don’t get is the sheer quantity needed. No matter how much money is thrown at new renewables projects we simply can’t build enough of them fast enough due to bottlenecks in supply chains, raw material mining, grid interconnection times, and other limits.

    New nuclear is the only other major option to speed up the transition away from fossil fuels. People resist it because of safety or waste concerns (neither are backed by data, nuclear is tied with solar for the safest tech, and it generates less radioactive waste than coal). Or they think nuclear has a big carbon footprint when you include the manufacturing and disposal (also not what the data says, nuclear is tied with wind for the lowest full lifecycle carbon emissions and is about half as much as solar). Or they argue renewables are cheaper which is at least mostly true, but it isn’t as clear cut either when you factor in the costs of connecting that many renewable power projects to the grid. Connecting one nuclear power plant to the grid is significantly cheaper than connecting the 100+ wind and solar farms needed for the same quantity of electricity. Not to mention the cost of storage.

    We want to be building renewables, but we can’t wait around for renewables to save us that’s just not going to happen fast enough, our best option is building as many renewables as possible and a bunch of new nuclear and anything else carbon free at the same time.

    1. We need to electrify everything that runs on fossil fuels. Cars, furnaces, industrial uses, everything needs to switch from burning oil, gas, and coal, to being electrically powered.

    But deciding what to electrify, when and in what order is complicated too…. adding to electricity demand before we’ve removed fossil fuel power generation from the grid, results in the scale-up of the fossil fuel generation to meet the increased demand. Until fossil fuels are gone from the electric grid, we should only electrify something if its efficiency is sufficient to still reduce emissions when we assume it’s powered by the most polluting form of electricity generation on the grid.

    Battery electric vehicles have reached that point including factoring in the high-carbon footprint of lithium-ion manufacturing. Even if charged exclusively with coal power a BEV has lower lifetime emissions than an ICE car. Even discarding ICE cars before their end of life to replace with a BEV will generally be a net win.

    Hydrogen fuel cell vehicles on the other hand (pretty much anything hydrogen-powered for that matter) aren’t even close. Using Hydrogen to power vehicles is not a tech we should be investing in right now.

    Even if you’ve built a dedicated solar or wind farm to power something you want to electrify that hasn’t reached that efficiency threshold, you need to ask if it’s better to use that solar farm to displace current coal or natural gas-based electricity generation than to power your newly electrified whatever. This is why even so-called “green hydrogen” is a counter-productive tech to be investing in right now.

    It’s also why some DAC and CCS techs shouldn’t be built yet. Even if you plan to build a dedicated solar or wind farm to power it. It’s often more impactful to just connect that solar/wind farm to the grid instead to reduce fossil fuel-based generation than to use it to power CCS. DAC and CCS is a rapidly developing space, we’re all hoping for some new breakthrough techs here that changes this story… so don’t criticize research in this area as a dead end… we don’t know that.

    Hopefully, you’re starting to understand why so many of these discussions are more nuanced than people on Reddit/Lemmy claim…. a lot of new electrification technologies are just on the borderline here for not causing more emissions, and it often depends on where you live and what will be scaled up to meet the added electric demand.

    All of this points back to why we need massive quantities of carbon-free electricity. Without clean electricity, these other techs aren’t a net win. Many things will cause a net increase in emissions if they’re electrified before carbon-free electricity is abundant. We need more new carbon-free electricity generation built in the next two decades than all the fossil fuel generation we’ve built in the last century put together. Even with ridiculously optimistic exponential growth projections of renewables, it is just not going to be enough. Until we’ve sequestered so much carbon that we’re back to pre-industrial levels, there will always be new techs that are “unlocked” by any additional carbon-free electricity generation.

    1. We need society to transition to lower consumption of everything in general. Every product or service you buy has a carbon footprint of some kind. There’s a LOT to be done around making smarter choices about what you buy, yes an EV is better than an ICE car, but public transit, electric scooters, bicycles, and ton of other things are better than any car, and not buying things at all if you if you don’t need them is better still.

    Capitilizim’s tendency to push towards ever more consumption is the largest driver of the problem here. We can’t have circular economies if the only metric we’re looking at is the bottom line. Our modern mentalities of disposable products, planned obsolescence, fast fashion, and other things we’ve come to associate with a “high quality of life” in wealthy nations need to be re-evaluated.

    1. We need better data to make better decisions. Corporations aren’t required to measure and report their emissions. We’re still largely making educated guesses at the carbon footprint of things because the only data available for most things are broad estimates and industry averages. Our supply chains are so interconnected, that trying to calculate how much of an impact a particular product has requires data from potentially thousands of companies that they’re not even collecting, let alone publishing.

    The EU is starting to mandate carbon reporting, but the US and Canada are lagging in this area. The US SEC proposed last year making reporting mandatory for publicly traded companies but caved to a bunch of pushback from corporations. They did pass a mandatory reporting rule a couple of months ago, but with significant retractions on what needs to be reported and how soon. They dropped a provision that would have required companies to report on emissions they’re causing to occur in their supply chains (known as “Scope 3” emissions), which would have put significant pressure on smaller and non-publicly traded companies to also report on emissions.

    Until the vast majority of corporations are tracking emissions, even the corporations that are trying to reduce emissions are limited in effectiveness because they are basing decisions only on how it impacts them directly and not what impact it might have elsewhere.

    Anyhow… that’s the “big things”….

    There are a lot of interesting little things that could be happening but aren’t, usually because they clash with a particular political ideology. For example, the government could pay contractors to go from house to house and upgrade the insulation, and it would have one of the best emission reductions for the dollar than almost anything we’ve quantified. But politically there’s a “It’s not fair to take money from my pocket to pay for someone else’s insulation” mentality that some people have that prevent many low-hanging fruit things…

    And on the flip side, some of the things that we’re doing that generally aren’t working include:

    Most carbon offsets on the market are bullshit, including a lot of nature-based offsets. The mentality of “don’t reduce just offset” emissions doesn’t work. I’m not saying there isn’t a place for offsets, there is, but the carbon offset market in general is full of bad actors. It’s trivially easy to misrepresent creative accounting as a carbon offset, even if it’s not intentional. And since there’s no tangible product delivered, some companies will sell the same carbon offset to multiple buyers. If you don’t believe me, I have a bridge carbon offsets to sell you.

    Another thing that isn’t working is most (if not all) RECs, GOs and similar market-based instruments for purchasing “green electricity” from the grid. You’re not changing the net emissions, you are literally just paying for the privilege of claiming your electricity consumption isn’t generating emissions. You’re not making more renewable get built, renewables are already cost-effective, they don’t need someone voluntarily paying extra for them for them to happen. If you pay extra for them, you’re just increasing someone’s profits.

    Note that RECs and GOs are not the same things as private PPAs, like when Amazon or Microsoft pay to build new nuclear to power their data centres.  Again lots of nuances here, but PPAs are causing additional carbon-free electricity to be built. RECs and GOs where you’re selling renewables that have already been built aren’t changing anything, just upping profit margins.


  • LOL, I work in climate science.

    Specifically in consequential carbon accounting analysis. Which is the branch that specializes in quantifying how much impact decisions and policies will have on greenhouse gas levels.

    We are fucked. We are so incredibly fucked.

    I comment regularly on social media about what actually needs to happen if we’re to limit the damage from WW3 to just seriously fucked. You can imagine how that goes.

    People advocate for things on Reddit or Lemmy about what we should be doing to avoid the disaster. Most of the time these things will have little benefit, and often will make things worse. I try to educate people but everybody has their pet issues usually based on whatever article they read last and they don’t actually want to seek the truth, just defend their opinion.

    It’s tough because they are all very nuanced issues, every decision has trade offs, makes things better in one way worse than another. People aren’t wrong about the small part they’re looking at, just its impact on the bigger picture.

    Everyone is pulling in different directions on this issue because the waters have been so incredibly muddied by the people who stand to lose from real climate action.


  • Yes you’re correct. I will qualify my previous statement as hydrogen powered road vehicles don’t make sense for now.

    The problem at the moment is that electricity generation is not carbon free and in most countries not even close.

    Unfortunately the transition to a carbon free electric grid is being significantly retarded by policymakers that are, as you say, myopic. As a result it will be at least two more decades before hydrogen makes sense.

    The carbon footprint of lithium battery manufacturing, is small compared to the carbon footprint of electricity generation. Until that changes significantly lithium batteries will continue to be a better choice than hydrogen fuel cell.

    Hydrogen may make sense in a future where we’ve eliminated all fossil fuel electricity generation and there’s an abundance of carbon free electricity that can be used to create green hydrogen as a form of energy storage. Though by the time that point comes, we may have developed battery technology or some other energy storage technology that doesn’t carry the same carbon footprint that lithium ion does today.


  • Hydrogen doesn’t make sense and never did as a strategy to reduce greenhouse gas emissions in vehicles.

    Most hydrogen is made from fossil fuels, and has a lot of emissions during manufacturing. But even green hydrogen, which is made by using carbon free generated electricity to split water into hydrogen and oxygen doesn’t make sense.

    If you’ve build new renewable power it’s more efficient to use it to charge batteries than to use it to generate hydrogen.

    There might be a case for compressed hydrogen, In vehicles where batteries are too heavy like aircraft.

    But for road vehicles, batteries are more effective at reducing emission.

    If you’re building any new renewable power, you’ll reduce more emissions by using it to displace coal power, the to generate green hydrogen.

    Some day when we’ve eliminated fossil fuel based electricity generation, Green hydrogen might start to make sense. But anybody trying to do it right now is not being as helpful as they could be.



  • I think I have you slightly beat… mine was an Apple II+, circa late 1981, with a disk drive, and a monochrome green screen monitor.

    First cell phone was around 1997. Though I honestly don’t remember what it was. I recall having a Nokia model from before they made that indestructible model in all the memes, as well as a Kyocera one that I could connect to a laptop and have wireless dial up internet at some abysmal speed like 20 kbps. (0.02 mbps). I had at least two more phones, including a Treo 650 “smartphone” before getting my first iPhone, a 3G. I’m on my sixth iPhone now.


  • The pro-lifers that think there should just be an exceptions when mother’s life is at risk need to also hear that the Texas law already has an exemption for when a doctor uses their “reasonable medical judgment” that the life of the mother is at risk or the pregnancy poses “a serious risk of substantial impairment of a major bodily function.”

    All pregnancies are a risk to the mother’s life, anytime a doctor performs an abortion in Texas they risk losing their license, and possibly even prison time. Which is why we have these case of mothers with unviable pregnancies that aren’t terminated until they’re almost dead.

    If you care about life, why on earth would you support a law that takes informed medical decisions away from doctors and their patients to put it in the hands of lawyers and lawmakers?


  • Who cares? Because I assure you, Microsoft doesn’t.

    20-25% of those webservers are running on Microsoft Azure hardware. Microsoft is the #2 cloud provider and has been slowly but closing their gap behind AWS in recent years. All of that is in large part due to them embracing Linux and open source support on their platform.

    Software isn’t the battleground, and hasn’t been for a decade. The people behind Apache and Nginx aren’t making bank on their web server dominance. Microsoft and AWS still rake in money hand over fist regardless of what software runs on their servers.

    The author of this article’s apparent attitude that this is some kind of indicator of Microsoft’s market failure is one of the most ridiculous conclusions I’ve heard in a while.


  • We’re both correct.

    LCOE is based on total operating costs of new electric power generation station over a 20+ year operation life. There are obviously a lot of assumptions in these sorts of analyses but Nat Gas is projected to become cheaper than Coal over the life of a new project, which some of that is expected to be due to carbon taxes.

    LCOE has some flaws as a comparable number when comparing wind and solar to fossil fuels, but is good for understanding what will be cheapest to build of fuel based generation.

    For current existing power stations, coal is cheapest of the fuels. The EIA numbers are here and here’s Statista research here on the historical cost of nat gas vs coal specifically which is frustratingly why coal phase outs have been so slow. Keeping existing coal plants operating is cheaper than building new almost anything.

    And you are correct, price is specific to geography and availability of each. My blanket statement of “coal being the cheapest fuel” is over generalized and not universally correct.


  • It’s expensive and dirty fuel.

    It’s definitely the dirtiest fuel by a good margin. But coal is actually the cheapest fuel. Which is the main reason it still gets used.

    Uranium used to be cheaper than coal, but now that we all but stopped building nuclear power plants it’s gone up significantly.

    Wind and solar are now cheaper than coal for electricity generation, if we can limit growth in energy demand to a rate lower than the growth in renewables, economics will eventually push all electricity generation off of coal.


  • I think you may have misunderstood friend. You’re not wrong and I’m not arguing against any of your points.

    A wind or solar farm is indeed much faster and cheaper to build than a nuclear power plant. Wind and solar farms take 8-18 months on average. Recent nuclear power plants have been taking 7-10 years.

    The nuance isn’t the time required for a single project, it’s the sheer number of renewable projects required that is the issue.

    I live in Canada, a single digit number of nuclear power plants here could replace all of the fossil fuel based electricity generation in our grid. That’s something that could be built within 10 years.

    We’d need ~1000 new wind and solar farms (not to mention storage) to do the same. We can’t make that happen within 10 years due to supply chain and grid interconnection bottlenecks limiting the number of concurrent projects we can do.

    I would ecstatically overjoyed to be proven wrong about this. But we need to get off fossil fuels as quickly as possible, and we can’t do that quickly with renewables alone.

    Frankly we’re fucked either way, but we’re less fucked if we build nuclear power in addition to as much renewable power as we possibly can make happen.


  • dgmib@lemmy.worldtoAsk Lemmy@lemmy.worldWhat gets you downvoted?
    link
    fedilink
    arrow-up
    4
    arrow-down
    1
    ·
    edit-2
    10 months ago

    You’re correct about renewables being cheaper… but faster is a more nuanced discussion.

    In the Canadian province I live in we generate 70% of our electricity with natural gas fired power plants. Roughly 20 TWh annually.

    To replace that 20 TWh/yr with solar power, we’d need to build ~150 more solar farms the same size as the largest solar farm in Canada. Plus enough storage to cover the grid at night or when the weather is cloudy.

    To replace that with nuclear power, we’d need 2 plants the same size as the smallest nuclear power plant in Ontario.

    The nuclear plants are significantly more expensive than the solar, that much is certain.

    But there are logistical limitations on how many new sources we can interconnect on the power grid in a given year. We simply can’t connect that much new renewables quickly.

    It doesn’t need to be a choice, we can do both renewables and nuclear. But if we want to get off of fossil fuels in the next decade, nuclear will get us there sooner.