Andr@ Gaili

𝗜𝘀 𝘁𝗵𝗲 𝗔𝗜 𝗕𝗼𝗼𝗺 𝗕𝘂𝗶𝗹𝘁 𝗼𝗻 𝗦𝗼𝗺𝗲𝗼𝗻𝗲 𝗘𝗹𝘀𝗲’𝘀 𝗟𝗼𝘀𝘀𝗲𝘀? Something about today’s AI economy doesn’t quite add up.On one side, we have neoclouds borrowing billions to buy GPUs, building expensive infrastructure, and paying for electricity, cooling, financing and operations - while aggressively competing to offer cheaper AI compute.Some are already struggling to turn enormous demand into sustainable profits.On the other side, we have thousands of AI startups.Some are creating genuinely revolutionary technologies. Others? An API, a nice interface, a few prompts, some clever marketing - and suddenly a multimillion-dollar valuation. 𝗧𝗵𝗲 𝗶𝗿𝗼𝗻𝘆 𝗶𝘀 𝗿𝗲𝗺𝗮𝗿𝗸𝗮𝗯𝗹𝗲. Those investing billions and taking enormous financial risks may end up making less money than those building relatively simple businesses on top of their infrastructure.And sometimes, the people creating those businesses dont even need to understand the technology behind them.Of course, cheaper AI is fantastic for innovation. But there is a difference between technology becoming cheaper through efficiency and services being sold at prices that cannot sustainably cover their full costs. 𝗪𝗵𝗮𝘁 𝗵𝗮𝗽𝗽𝗲𝗻𝘀 𝘄𝗵𝗲𝗻 𝘁𝗵𝗲 𝗯𝗶𝗹𝗹 𝗳𝗶𝗻𝗮𝗹𝗹𝘆 𝗰𝗼𝗺𝗲𝘀 𝗱𝘂𝗲? What happens when GPU financing, depreciation, electricity, infrastructure and the cost of capital must all be reflected in AI pricing?How many of todays multimillion-dollar AI startups would still have viable business models?And heres another uncomfortable possibility: if heavily indebted neoclouds fail, their GPUs dont disappear. They may return to the market at distressed prices, pushing compute prices even lower.More pressure. More losses. More consolidation.Meanwhile, the next AI startup raises another few million. 𝗪𝗲 𝗺𝗮𝘆 𝗯𝗲 𝘄𝗶𝘁𝗻𝗲𝘀𝘀𝗶𝗻𝗴 𝗼𝗻𝗲 𝗼𝗳 𝘁𝗵𝗲 𝘀𝘁𝗿𝗮𝗻𝗴𝗲𝘀𝘁 𝗲𝗰𝗼𝗻𝗼𝗺𝗶𝗰 𝗲𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁𝘀 𝗶𝗻 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗵𝗶𝘀𝘁𝗼𝗿𝘆. Billions invested in infrastructure so that others can build million-dollar businesses on top of potentially unsustainably cheap compute.AI will undoubtedly transform the world.But transforming the world and making money are two very different things. 𝗦𝗼 𝘄𝗵𝗼 𝘄𝗶𝗹𝗹 𝘂𝗹𝘁𝗶𝗺𝗮𝘁𝗲𝗹𝘆 𝗰𝗮𝗽𝘁𝘂𝗿𝗲 𝘁𝗵𝗲 𝘃𝗮𝗹𝘂𝗲 - 𝗮𝗻𝗱 𝘄𝗵𝗼 𝘄𝗶𝗹𝗹 𝗯𝗲 𝗹𝗲𝗳𝘁 𝗽𝗮𝘆𝗶𝗻𝗴 𝘁𝗵𝗲 𝗯𝗶𝗹𝗹? #AI #ArtificialIntelligence #Neocloud #GPU #AIInfrastructure #DataCenters #AIStartups #AIEconomics #TechInvestment

Mums vajag 50%!!!

Nav saprotams ka laba strada @ltvzinas … Kapec apzinati jaslepj informacija???

The supposed promise that NATO would never expand is a myth. Every sovereign country has the right to choose its own alliances. One dictator’s wishes do not create obligations for the rest of us. He should put his own country in order before trying to dictate the future of others. #NATO #Ukraine #Latvia #Estonia #Lithuania #Poland #Finland #Sweden #BalticStates #EuropeanSecurity #Sovereignty #StandWithUkraine

Jalikvide!

Pēc šāda zinatnes un Latvijas apkaunojuma tā SIA LZA naftalinēka ir jalikvide!

Problēma ar pilsonības piešķiršanu šādā veidā jau nav pašā pilsonībā kā tādā, bet gan tajā, ka visu šo gudrinieku dēļ, kuri bīda šādus jautājumus, neveic pienācīgas pārbaudes, lobē savas intereses un pieņem šādus lēmumus, mūsu valsts var kļūt par apsmieklu. Turklāt mūsu partneri arī vēro, kā mēs spējam aizsargāt savu valsti un tās intereses. Tas nav tikai reputācijas jautājums - tas ir reāls drošības apdraudējums (real threat), ja tiek pieļauts, ka sistēmu var tik viegli apiet un izmantot savās interesēs. Var pienākt situācija, kad citi varēs teikt: „Redz, kā mēs viņus apčakarējām!”Un te nav runa tikai par viltus zinātniekiem vai akadēmiju, kuras reputācija jau tā ir zem nulles. Runa ir par kopainu - par sistēmu, kuru iespējams apiet, manipulēt un izmantot savās interesēs. Problēma ir ne tikai konkrētajos cilvēkos, bet arī tajā, ka šāda sistēma vispār pieļauj šādu situāciju rašanos.

Kautkā nes pēc naftalīna :). Un es ne par vecumu, bet par valsti, kas aplipusi ar dēlēm... delfi.lv/193/politics/120139…

Visas nozares finansejumu un klientus mekle pašas! Lūk arī lai šie skaistuļi dara to pašu… delfi.lv/kultura/52862727/sc…