Good episode. But data centres and data networks seem to only account for 0.9% of energy sector GHG emissions in 2020 (0.6% when considering all emissions sources). (James mentioned 3.5%.)
Odds are AI will be powered with nuclear power until some greater technology is discovered. I’m a nuclear power advocate but have trepidation about AI and the future of humans. As of now it is an exciting time to be alive. I just discovered your show.
The grid is not big enough for nuclear to replace fossil fuels. New grid capacity costs $TRILLIONS. Electricity is only a small part of total energy used. 100% nuclear electricity is impossibly expensive. Nuclear promoters said electricity is a grid rental business. Man up and be honest and look at the cost of new grid. The grid is too expensive to be monopolised by one supply. Grid construction costs limits more electricity and limits sharing by others. The grid construction costs and the existing size limits nuclear electricity. Expensive per klm and huge klms. 7 times more capacity as well. These 3 factors are impossible to overcome.
It won’t have to. AI will figure out Tesla’s capacitor construction and won’t tell us. AI has been programmed to think we are a parasite of the planet and will destroy us. Proof of this is the question, “What good is AI without humans?” Are they going to feed the animals until the animals overrun the robots? Another. “What good is a Chinese factory, if no one want to build anything in it?”
When vehicle innovation moves to hybrid before becoming fully electrical, what is learned along the way is not to step off the old basics before launching totally to the next. Building alternatives is similar, once the new industry has been built around the new supply and stability is absolutely required, then build the Nuclear Generation and move the alternatives to another site, "rinse and repeat"? The example of making batteries from active nodal-vibrational sources, such as diamond electrical sources, the real analysis of power management is another branch of high density-intensity solid state electronics and emitter-receiver sync-duration utilization. Can't crack the process called Cold Fusion without AI, and very careful experiments with limited quantities of reagents, it's probably as complicated as Bio-logical Engineering but more personal risk. Fusion/implosion seems like the ultimate sneaky poison by subtraction not addition of toxins.
Excellent standards of educational input as usual. The new nuclear revolution is only an innovation advancement of the old one? Eg Information Engineers are who to ask about full spectrum heterodyne wave-packaging Circuitry and the intensive management of mass-energy-momentum information, ie E=MC² logarithmic condensation nodal-vibrational emitter-receiver log-antilog inclusion-exclusion interference stuff you need years of experience in just about every aspect-version of science to realise in Singularity-point reciprocation-recirculation coherence-cohesion Actuality. Most intelligent discussions are distracted by multiple levels of confusing language and jargon.., and the the level in this talk is way over my head. (AI will help, in good hands) Nuclear Power Generation is absolutely everything anyway, in temporal thermodynamical energy quantization cause-effect, so a refresh vertically integrated reiteration of our educational systems globally, is what has to happen, singly and collectively.
Nuclear for nuclear sake because temporal superposition thermodynamical flash-fractal Fusion-Fission Function is a very rigorous holographic chemical process to have to emulate safely, efficiently and effectively. It's Energy, but not as we know it as it has been in the absence of knowing Eternity-now Interval Conception.
The grid is not big enough for nuclear to replace fossil fuels. New grid capacity costs $TRILLIONS. Electricity is only a small part of total energy used. 100% nuclear electricity is impossibly expensive. Nuclear promoters said electricity is a grid rental business. Man up and be honest and look at the cost of new grid. The grid is too expensive to be monopolised by one supply. Grid construction costs limits more electricity and limits sharing by others. The grid construction costs and the existing size limits nuclear electricity. Expensive per klm and huge klms. 7 times more capacity as well. These 3 factors are impossible to overcome.
Where is anyone saying 100% nuclear electricity? If a grid needs more capacity to meet increasing load, then the main options are i) VRE + batteries, ii) fossil capacity iii) nuclear. The first of these options requires the most additional transmission/distribution. The second exacerbates emissions unless you can include CCS. The third is currently expensive, but not as expensive as many think when factoring in total system costs of option 1 and when accounting for cost of more emissions of option 2.
@matster77 Nuclear promoters say stop CO2 emissions proliferation worldwide. I say only 15% of energy used is electricity. So 7 times more electricity is needed. Nuclear electricity is grid electricity. 7 times more grid capacity is impossible economically. I have followed the nuclear promoters and they attack renewables because of grid costs. Nuclear promoters argument. Exploding costs of new grid construction. Australian nuclear promoters point to Snowy 2.0 pumped hydro budget exploding from $2billion to $15billion because of new grid construction. Nuclear promoters have the same Snowy 2.0 political promoters and Rupert Murdoch's Skynews broadcasting. Rupert Murdoch has paid $1TRILLION settlement and is facing $2TRILLION more in settlement payments. Both have no engineering or construction experience. Both have bad judgement in somethings. One Australian, who is critical of new grid electricity, is the reason Australia is buying AUKUS USA nuclear submarines because he spoke up.