By $AMZN own admission (as did $MSFT), the rate of depreciation of $NVDA GPUs is far faster than prior to Ai. “The shorter useful lives are due to the increased pace of technology development, particularly in the area of artificial intelligence and machine learning.” They’re basically junking old H100 and H200 in warehouses and replacing them with new generations due to power restraints. Ie rotating in Vera Rubin to replace older generations that had a real life of 2-3 years. The worst part is, the paper depreciation is over 6 years, meaning that even while many GPUs were actually physically depreciated in 2-3 years, the losses from that still haven’t shown up on the balance sheet even while not generating revenue; those losses are being spread out over an additional3-4 years. Now $AMZN just saw its operating income peak, and it’s about to reverse in a very way with its recent exploding Capex spending set for depreciation… Amazon saw depreciation rise to $9.2 billion vs $5.6 billion YOY for Q3. A 64% increase. Next quarter comes the BIG depreciation costs to hit operating income (reversal), with true negative operating income to hit in late 2028.

Less than a month ago, OpenAI was looking to raise “at least” $30 billion at a valuation of $1.4 trillion. Today it raised $11 billion at $1.25 trillion valuation and faced “tepid” interest.

My friend made a really impressive app. It's live and has users. He has no idea where it's hosted.

I don't see HCI as CS research at all. It's using science, and it's using computers, but CS is a math field, not science.

3SAT in subexponential time (2^{n / log^eps(n)}) is maybe the biggest surprise I have seen during my 20 years following theoretical computer science (still remember the feeling 25 years ago seeing the diff a 3 versus a 2 can make for complexity 😅)

a big deal if correct.