During the mid 2010s, I made a supposition that most unprofitable VC backed companies were spending $.40 of every $1 raised on FB and Google ads and AWS compute. It turned out to be largely right.
Unfortunately, we are back to this same cycle in AI with NVDA but I worry that it’s now $0.60-75 of every $1.
Every entrepreneur should be aggressively trying to find a way to shrink their compute OPEX and use a managed service that delivers production quality speed. Otherwise this is all just sandboxing, toy apps and means little of anything.
Boards should be asking for OPEX breakdowns and pushing their companies to look at every possible way to spend as little as possible for compute.
@DragonflyTea top tip print the name of tea on the tea tag.
Can’t figure out which one is traditional English, Moroccan Mint or Cape Malay Chai ☕️ @SoVeryBritish problems.
Grow London Global: Cohort 2 - let's do this! 🚀
Today we're launching our second cohort of London's fastest-growing scaleups - 100 companies about to go global.
See who's on our cohort here: https://t.co/7WVeedQeyi
Quick speed test after a week with @CommunityFibre so far so good. Helps when you are the only one in the street but probably not for long. Toasty hot fibre speeds!