Ed Zitron draws conclusions from Microsoft's downscale of future data centers. Ed interprets this as MS losing faith in the future of AI but it could be a turn away from brute-force model size expansion and toward DeepSeek-style efficiency improvements.
I've been waiting for "Lottery Ticket Hypothesis" breakthroughs and I hope this is the beginning.
Ed Zitron draws conclusions from Microsoft's downscale of future data centers. Ed interprets this as MS losing faith in the future of AI but it could be a turn away from brute-force model size expansion and toward DeepSeek-style efficiency improvements.
I've been waiting for "Lottery Ticket Hypothesis" breakthroughs and I hope this is the beginning.