A call is only graded once it's had time to play out and the stock has moved beyond its normal range — long-term views get a full quarter before we'll say they're tracking against.
Gavin Baker argues OpenAI's competitive position is at risk because its largest training cluster is still smaller than xAI's Colossus, and significant talent departures (notably Mira Murati resigning during a fundraise) signal internal problems.
The best information I have is the largest cluster Microsoft has after panicking is still smaller than xAI's cluster in Memphis... I do think a lot of talent has left OpenAI. I thought it was a really shocking statement from Mira Murati” ⚑
Gavin Baker argues xAI's Colossus supercomputer (100K+ coherent GPUs, the world's largest) is a breakthrough that no one else achieved, and if scaling laws hold, Grok 3 should take the lead in AI model capability in early 2025; his firm is also an investor in xAI.
Grok 3 should take the lead if scaling laws hold in January or February... My firm is an investor in FCI.” ⚑
Gavin Baker argues xAI's Colossus build effectively saved Nvidia from a difficult 6-month period when Blackwell was delayed, as it prompted others to rush out and buy Hopper GPUs; and with Blackwell arriving next with new architecture and networking, Nvidia's position remains strong.
You can argue that Elon doing that in a lot of ways kind of saved NVIDIA from a tough 6-month period when Blackwell was delayed, because everyone who was waiting for Blackwell... rushed out and bought a lot of hoppers to try and do it” ⚑
Gavin Baker cites Alphabet/Google as one of the clearest examples of positive ROI on AI investment, with return on invested capital going vertical since they ramped GPU CapEx.
Meta and Google have shown the best ROIs on AI.” ⚑
Friedberg suggests Bitcoin will take over gold as a store of value, implying gold loses its dominance.
It should be kind of store of value and alternative. It's just going to take over gold.” ⚑
Gavin Baker warns that MicroStrategy's debt-funded Bitcoin accumulation strategy will become unsustainable as the underlying business ($400M revenue) is too small to support growing interest expenses unless fixed income markets fully accept Bitcoin as collateral, which they don't yet.
I think a lot of hedge funds are short MicroStrategy, but I have no horse in the race. But his, the underlying business that pays the interest expense on the debt only does $400 million a year in revenue and it's high gross margin revenue.” ⚑
Joe Lonsdale, who has co-founded and invested in multiple defense tech unicorns, is highly bullish on the new wave of defense tech primes like Anduril, expecting 7-10 new primes to emerge and generate huge returns for investors who can access them.
It's going to be like 7 to 10 new primes. One of them's obviously under all, I think one is probably Seronic and Everest, but we'll see. But like there's going to be 7 to 10 new primes and that's what it's going to be. So there's going to b” ⚑