GL26

comment section looks like it's filled with bots, a lot of commenters looking like they are giving their opinion as if they had read the whole pages, sometimes irrelevant

hot take: software IS getting better. You give access to SWE to more people = more peers to review and test code/algorithms/apps... the fact that OSes had these downsides was technologic immaturity (just like trains that ran of coal steam engines in the XIXth century vs now that are electric). The more mature a tech is, the more everyone gets to enjoy it's outputs, and the more it grows.

Two sides : one side defending that cognitive decay is inevitable, we will all be jobless in 5 years, starting with white collar jobs, and then blue collar ones. Detroit become human scenario Other scenario : I believe the notion of the "difficulty" of a problem just rised really fucking higher then 6 months ago, humanity will be able to tackle problems of a whole different scale. The bottleneck : financing this revolution is going to be really difficult, you cannot borrow and lend money indefinitely, someone has to pay for R&D expenses. Moreover, the real question is whether this can transform the global economy in a durable and robust way, for now if datacenters decide to shut down, or 20% of all gpus stop working, we are in deep trouble :)

maybe stupid question but how is it possible to not be protected against supercrawlers like this, when you own large amounts of data, specially if all requests are coming from the same server ? Aren't cyberprotection services like CloudFlare work against this kind of "attacks" ?

the bottleneck is now how fast can you plan requirements, how fast can you check those requirements were met. Stack : Greptile + TesterArmy

hot personal take: this is not esthetic

someone's getting auction-poached at openai and anthropic

Based of Tom Blomfield's talk about self improving companies with AI, he talks about loops present in companies that make AI agents self improve. He talks about loops that look like this : - Data acquisition / feeding mechanism - Policy layer - Tools - Quality Gates - Learning Mechanism

The policy layer is the one that looks the most abstract to me, what does this look like ? Is this a CLAUDE.md ? Is this a state action graph like what you learn when you start RL ? What does this look like ?