@eatonphil Nooo, is it really an issue? Lichess is so much better everywhere. Free analysis, opening and no ads, in contrast with chesscom.
I don't use move confirmation, so maybe some report to their repo would be nice!
@codependent_ai I'm tired of no one talking about this. It's such an important topic, yet most devs assume we can put agents running in parallel. With companies enforcing more limits, it's impossible now.
@hybridllm@simonw I think we are in for a very different future. Highly paywalled access to things, and security only for those who can pay millions in tokens. For us mere mortals, forget it.
@fatih No you don't, but people look up to top software developers like you. If all they read is "I don't type anymore my use my agents do it for me" it makes others feel they have to do the same, thereby not gaining the experience and intuition you did. Sorry if it came across harsh
@badlogicgames I think this is also true as long as you can keep using agents indefinitely. Without it, there is no human way to review this.
However, if you are oai or anthropic then you can afford all million of tokens. For rest mortals, we are not there yet. Humans and quality are needed.
@reet416 And are you expected to contribute something just because they switched? That's terrible. If you hear people talking and you join, you are not expected to say anything.
That they switched is not the expats fault. I think its not so easy to figure out what's best.
@iwooky@thorstenball They all have their quirks and different things, but mostly is just that it's expensive. Amp does not subsidize your tokens, so by using AMP you are acknowledging that your token usage will exceed what others charge you.
It's great but I can't afford it.
@georgemillo Because agents like amp don't subsidize it , so you spend a lot. Additionally, some people fear they are missing out on writing slop that they pay multiple flat 200€ subscription to avoid limits.
@__tosh@badlogicgames I don't think is per se necessarily about cleaning up. If there is a leak, who is to blame then? So then the person that pushed those commits or the manager behind the agente would be accountable for something that wasn't checked properly
@ryancarson Do you think sourcegraph will make their own models?
I don't know how relying on big companies for models can be good long term,but of course I have no clue
@eatonphil Refreshing to have a different take from a professional that is not "AI can code faster and better than you, so just give up"
I just wished strengthening your skills wasn't seen nowadays as a "why bother". Thx