@qofenoai@heydamir@8008s137 Sometimes it is, but every word in the SKILL.md effects its results. LLMs often put filler and repetition when they make skills
@grok@theo@grok are you sure you can combine Devin and T3 Code? Also search for any benchmarks comparing the three harnesses or comparing these to Codex and Claude Code.
@cognition This benchmark is not at all useful for @cognition users because the models are evaluated across 6 different harnesses. You’re a harness company. We want to know how they perform on your harness (and ideally how native harness compares to Devin)
@kaseyklimes Obviously Ghodsi thinks the context needs to be as narrowly focused as possible. You can’t have one agent be an entire company. Agents need a defined role and limited toolset to only see the relevant knowledge
The bottleneck of AI adoption is not frontier model quality or compute. It’s context engineering (and corporate policy/trust). Localized agents need knowledge, tools, structure, and purpose; and that’s not easy to provide. That’s where talented human engineers will shine
we will need to rebuild and rewrite much of software to adapt to a new, faster pace.
code used to arrive one drop at a time. soon, if not already, code will be written as if it were a continuous stream, like a violent river twisting and turning as it pleases.
this is not a new thing: big tech cos have had to deal with this with the enormous scale of their engineering teams. but now ai has brought an incredibly elastic and capable workforce wielded by every engineer. this is your problem now, and the only way to adapt is to evolve your systems so agents - even and especially the ones with little context about the right way of doing things - can thrive. make the easy thing the right thing.
your codebase is becoming organic architecture. agent contributions grow around the constraints and guardrails you’ve designed. good software engineering matters even more now! and no, the continuous stream of code is not something you have to watch helplessly. be more ambitious and design systems, frameworks, and languages to control the chaos.
the time for mourning coding is long over
@nikillinit Hopefully you will outlive terrible microservice design. It should die in a couple decades. That said I’ve never heard of you before and this video is amazing
@SyntheticBeef Most current software was developed before LLMs so now is the best part. Dissecting software that looks like this and completely replacing it
@SyntheticBeef Most current software was developed before LLMs so now is the best part. Dissecting software that looks like this and completely replacing it
@Fraudmmich@RileyxxHuey@clashreport I mean if you go back to the German Confederation that includes Czechia, Luxembourg, Slovenia, half of Poland and half of Slovakia
@jcbtorti@RexTheodericu@Sollberg1@clashreport “dialect” and “language” are so loosely defined in many cases that you could say Dutch is a dialect of German
And don’t forget about Pennsylvania Dutch, the German dialect spoken by Amish Americans
@Sollberg1@clashreport No but Melechon is denying the ethnic identity of the Germans and French and instead saying French identity comes from the state. The whole point of my post is criticizing Melechon’s quote that my post starts with
@Sollberg1@clashreport German dialects are mutually intelligible *today*. But to say German identity was created from the language and not the ethnic group is nonsense