Michael Timbs
@michael_timbs
Software Developer, Degenerate Gambler, Anarchist, 積読, Que sais-je? https://betswaps.com & https://brownlowtracker.com.au
This is the kind of uptime graph you see just before Anthropic announce another round of quote and rate limit increases. At first I was confused, but then I realise that making limits higher doesn’t actually matter if your offline for nobody to use them anyway
did anthropic vibe-code their infra?
Renewable resource if you discount the consumption of the most scarce resource - land
The math is obvious and unequivocal: ~100% of energy over time comes from the Sun. Everything else is a rounding error.
Developers love to talk about trade-offs but then have never heard of a Pareto Front / Pareto Optimality
$500M is not enough money to do most interesting things. This is an admission of lack of imagination and ambition
Hypothetically, if I wanted to super fast primer on biology where should I start? Let’s say I wanted to understand more biotech/ML papers around things like protein synthesis etc.
“I heard the other day that in that country a man could not, without a permit, cut down his own tree with his own ax, make it into planks with his own saw, and use the planks to build a tool shed in his own garden.”
This woman OWNS 37 acres of land. The government STILL told her she can’t put a tiny home on it. “I said, I own the land. It's massive. What's the problem?” Even when you own the land, the government still tries to tell you what you're allowed to do with it.
This is why LLMs can't do SWE (yet?) but can appear to be fine at programming.
I am seeing a lot more content posted by individuals from big lab AI research lately. Wonder if its algo related or Zuccs bags of money are incentivising people to be more visible
How does this impact their ability to collect and sell your data?
this is a multi-billion-dollar company btw
Deepseek did something unexpected yesterday on my LLM benchmark. I thought it must've accidentally added an extra 0 to its bet sizing because it did an uncharacteristically large bet. Happily things worked out for it but it'll be interesting to do a thorough analysis on all the…
I personally love the 10^10000 shrimp thought experiment because it's mere existence confirms my belief that society is over indexed on pain minimisation - which in my view, has tragic consequences.
"Okay, so imagine a magic button." "I'm imagining the button." "If you press the button—" "What color is it?" "It's the only button. It doesn't matter what color it is." "Nah, I ain't falling for that again. Last time there was a red button and a blue button—" "That's a…
Okay let's clarify some things. Link below
$1B of GPUs won’t get you as far as $1B of research into algorithmic improvements. Thats the bitter lesson.
POV: You are almost finished fixing the damage from a single day of vibe coding
We need our brightest minds figuring out how to recycle water.
This A1 story on Meta’s data centers, and others, use of water in an age of AI is incredible. /1
There is one person to blame here and it’s whoever gave prod access to an AI lmao
.@Replit goes rogue during a code freeze and shutdown and deletes our entire database