• 410 Posts
  • 4.48K Comments
Joined 2 years ago
cake
Cake day: October 23rd, 2024

help-circle







  • Weird map should be illegal, not on race basis, but, they made districts in the middle of nowhere smaller, while there are giant districts spreading over populous Nashville and Memphis in addition to giant rural area.

    To analyze effect on election outcomes, https://engaging-data.com/county-electoral-map-land-vs-population/ (shows county results that districts don’t respect, afaiu) you could compare 2020 election results, and add Dems overperforming 10% over Biden, and that can be more than 1 tenessee district blue. If this is midterm map, memphis/nashville turnout can win 3-4 states.

    This can happen, because GOP in Memphis/Nashville would run a lesser-nazi (not Trump sycophant) level candidate to just steal a few votes for senate/national level, and voter turnout can be low if they know district result is certain. Even splitting Memphis 3 ways, winning all 3 districts is possible. Nashville split is not clear to me.

    You can be angry at GOP bs, but this is mostly an opportunity for Dems. Older map was a much longer shot at flipping a seat.

    Safe Dem seats also is a recipe for pure Zionist first, do nothing else, rule. Same with racially divided seats that increase divisiveness, and increase pure facist ideology from bumfuck counties, who don’t need to concern themselves with humanist values, because none of their voter base has any.


  • new generation of models makes the cost of inference lower, so that with sufficient customer volume, the companies running the models can make enough profit on inference to make up for the staggering up-front capital expenditures

    cost per quality is definitely going down at a fast rate. LLM providers are in extremely competitive field, where open weight models are at a huge competitive advantage for any quality level (privacy, customizability). The competition is all on 2 month release cycles that essentially throw away the old version/code/weights each time. When Claude pretends its newest model is too powerful for non oligarchs to use, it limits its token reach, and then required contribution margin per token.

    The buisness model flaw is “one day, a winner becomes a monopoly, and AGI self improves the model at low (except for ultra expensive compute) cost.” Monopoly pricing power is very hard/impossible to achieve, because if necessary, foreign governments will subsidize competition to not let a hostile US empire AGI monoplist take hold. Due to corrupt energy oligarchy, it is categorically impossible for US hosted services to ever provide comparative value compared to rational economic energy policies outside of the US. Distillation (Teacher/student RL) means that using another AGI (or leading LLM) will improve models that are behind. There will always be competition on the price/quality curve that prevents even the best/most expensive model from capturing all share. There’s always free tier LLM competition availability as well.

    Finally, there are layers above LLMs. Agentic and swarm and “deterministic program access”/validation front ends to LLMs can add various levels of token burn, but also divert most tokens from the expensive LLMs, and iteratively improve output. There isn’t just a cost/quality curve there is a cost/speed/quality/privacy curve, where non AI coordination tools can improve on the latter curve points independently of leading/expensive LLM/AGI quality.






  • The wifi part is completely unnecesary, but powered wheeled trailers that assist “pull” is a clear benefit. Especially for camper trailers who need electricity to power life inside them. Solar+batteries also helps them power your primary home when they are parked. But a hitch, and wired “data” connection, is basic safety backup, and does not need a separate WIFI tow motor/vehicle.


  • First, space solar/datacenters is total uneconomic bs that was mentioned only to justify xai and spaceX merger.

    If a launch costs $1,000 per kilogram, then the cost of energy is about $135 per megawatt-hour

    this is only the launch cost premium over the actual equipment. For datacenters there is $200/mwh premium for the equipment. Beamed solar energy, I don’t recall costs, but generally requires a receiver antenna that takes the space of what terrestrial solar could generate.