slopusila

move to SF. that's the place AI will nuke first

it turns out that all those jokes about EU regulating the curvature of the cucumber were on to something

another cookie warning disaster incoming

hopefully AI will wake them up and save us from all this nonsense

now you are debating exactly how GPT should be credited. idk, I'm sure the field will make up some guidance

for this particular paper it seems the humans were stuck, and only AI thinking unblocked them

after those 30 min you can manually ask it again to continue working on the problem

it's called ethics and research integrity. not crediting GPT would be a form of misrepresentation

hey, GPT, solve this tough conjecture I've read about on Quanta. make no mistakes

yes https://www.anthropic.com/research/end-subset-conversations

what is the business case for a text editor in a code writing agent world?

maybe they could pivot into the luxury boutique hand-crafted artisanal code market

what if I prompt it with a task that takes one year to implement? Will it then have agency for a whole year?

from the creator of openclaw - a lot of websites block/rate-limit non-residential IPs

driving a browser in the cloud is also a bit of work

but you could put a proxy on your residential machine

more tokens, less reliable, dont work in all CLI agent harnesses

agents can understand them. and you can view them in the terminal

seems like it requires an API key to your proprietary Ensue memory system

Most likely a Nick Land ref:

This is because what appears to humanity as the history of capitalism is an invasion from the future by an artificial intelligent space that must assemble itself entirely from its enemy's resources.

My concerns about obsolescence have shifted toward curiosity about what remains to be built. The accidental complexity of coding is plummeting, but the essential complexity remains. The abstraction is rising again, to tame problems we haven't yet named.

what if AI is better at tackling essential complexity too?

you can invent a new language, ask LLM to translate existing code bases into it, then train on that

Just like AlphaZero ditched human Go matches and trained on synthetic ones, and got better this way

I know another version, where 27 was a joke about the guards and then they come and beat him

including email addresses, phone numbers, and other unspecified “internal metadata.”

Substack specified that more sensitive data, such as credit card numbers, passwords, and other financial information, was unaffected.

I hate it when companies do this.

passwords and credit card numbers are easily changed.

names, emails and phone numbers are not.

how about support for "auth import-token" so you dont need to keep SLACK_TOKEN in env

two engineers implementing the same task are not-deterministic

yet nobody complained about this

in fact engineers appreciate that, "we are not replaceable code monkeys cogs in the machine as management would like"

even this is an intermediary stage

as elon musk said, the next phase is 100% AI run companies, any sort of a human in a loop, even if in a minimal role, will collapse the productivity

elon musk saud that the next wave is agents (optimus human in a box) clicking around the old SaaS apps, not agents writing new apps to replace them

microsoft ceo said the same thing - agents will be using windows/office and be the bigger buyers of licenses

replacing SaaS software at this scale is hard, nobody will risk it, at least near term

what would be more massive would be phones not booting up because of a botched update. this way you can just switch back to the old partition

that's how some phone OSes update the system (by having 2 read only fs)

that's how Chrome updates itself, but without the symlink part

on prem economics dont work because you can't batch requests. unless you are able to run 100 agents at the same time all the time

most of those subscriptions go unused. I barely use 10% of mine

so my unused tokens compensate for the few heavy users

why do you care about uncommitted virtual memory? that's practically infinite

after the models get good enough to replace coders they will be able to start increasing the subscriptions back up