I'm still kicking myself because when I first heard about Bitcoin (in 2010) I was working on a research project that required three GPU clusters (each costing between $15k and $20k). The contract for the project was delayed and I seriously thought about "benchmarking" the hardware by Bitcoin mining until work on the project could continue. I never actually did any mining and the hardware sat idle in our lab for a couple of months. I don't have the stomach to calculate what the Bitcoins that hardware was capable of generating would be worth today.

Bitcoin has had the ability to make nearly everyone look foolish at one point or another. Try not to kick yourself too hard over it.

Exactly. A lot of people sold all of their bitcoins when it reached insanely high rate of $1, and those who bought at $1 probably sold them at $10.

It wasn't until over a year later (July 2010) that the first block was likely mined using GPUs.

Satoshi implored people in December 2009 (several months after this post) to postpone mining using GPUs for as long as possible[1].

[1]: https://bitcointalk.org/index.php?topic=12.msg54#msg54

Actually the first person known to GPU-mine (supposedly an Apple OS engineer, friend of "BitcoinEXpress" on bitcointalk.org) claims he did it in May 2010, wrote his own miner, running on 3 x AMD Radeon HD 4850 graphics cards. The claim sounds plausible. There was a small bump in the global hashrate in May that seems to correspond to the power of this setup.

I commented on it here: https://bitcointalk.org/index.php?topic=40876.msg520645#msg5...

Satoshi's pleading to not port bitcoin to big endian machines is shocking. Hopefully that has been fixed!