• 1 Post
  • 340 Comments
Joined 2 years ago
cake
Cake day: July 11th, 2023

help-circle

  • Not intentionally being misleading, as I do have a footnote calling out not including carbon emissions from electricity generation as they vary so wildly based on the energy source.

    But unlike ICE cars, EV emissions from energy sources are improving over time as nations build more and more renewable energy sources. Your linked report is correct, but potentially out of date already - the UK for instance was already at 58% in 2024, with a goal of full of 95%+ by the end of this decade.

    Here in Australia, our uptake of residential solar has been so high that our energy providers are offering free electricity during peak daylight hours to all customers to help use up all of that excess production. It’s quite feasible for a significant portion of us here to be able to not only recharge an EV for free, but with next-to-no CO2 emissions.

    Additionally, we now have a big Government subsidy in place to install batteries in our homes as well: ~£4,000 for a ~30kWh system, fully installed!

    I share your love for older cars, but with a toddler and another one planned - we need to have a modern, safe car for peace of mind. But believe me, I will be ensuring that I disable as much telemetry as possible due to privacy concerns.

    But for a secondary/weekend car - there is always the option of electrifying an older car, allowing for the best of both worlds - in a sense!



  • Rough math involved: production of a new EV results in between 8-15 tonnes of CO2 emissions, depending on the size of the batteries and vehicle trim.

    But let’s aim for somewhere in the middle and take ~12 tonnes as a yardstick.

    ~12 tonnes of CO2 emissions equates to roughly 1,350 gallons of fuel.

    Depending of fuel efficiency, this would equate to between 20k~45k miles.

    Feel free to double-check my math in case I did anything wrong, but it does validate that most of these „facts” around EVs are likely FUD spread by fossil fuel aligned sources.

    ETA: initially forgot to include CO2 emissions from electricity generation - but this varies wildly based on source (nuclear, hydro & renewables at 0 etc.)


  • IP/trademarks/copyrights/etc.

    This is likely going to be the main reason for the takedown notices, Sony will be exercising their legal rights in order to defend their trademarks & copyrights on Concord assets.

    If a company doesn’t defend them vigorously, then any unlicensed works that are allowed to exist are then used as legal precedent moving forward to null/void such copyrights and trademarks.

    As an aside, Sony is a global corporation and can likely choose to write down these losses in the most preferred region to maximise the tax offset - so likely either the US, or Ireland.





  • The best way to think of them is as cousins; they are similar - but not exactly the same.

    They focus more on higher VRAM and CUDA cores compared to GPUs, while forgoing 3d acceleration capabilities.

    But they both come out of the same factories; so when the demand for AI cards is as high as it is now - and Nvidia can sell as many as it produces with a higher margin than GPUs, there is little incentive for them to produce more GPUs and sell them at a competitive price.

    So when the AI bubble bursts, demand for AI cards will crater - and there will be no financial incentive to mass produce them in such high quantities. This frees up production capacity at the TSMC factories, incentivising production of lower margin products like GPUs.

    Economics is largely a game of supply & demand; when supply outstrips demand, prices fall as sellers search for buyers. When demand outstrips supply prices go up as buyers search for sellers.



  • Assuming the AI bubble bursts before then, we might actually see somewhat reasonable pricing for next-gen consoles.

    A major reason why prices have remained so inflated for so long post-COVID is because data centres have been sucking up every bit of silicon that TSMC has been able to pump out for both Nvidia and AMD.

    But that would be honestly a very small upside, compared to what would likely be the Mother of All Stockmarket Crashes. The market cap of the Top 10 AI-related stocks is greater than the current US national debt, they aren’t in a position to be able to reasonably bail out those companies when it all eventually goes to shit, like they do in 2008.