At first, we kind hated AI tools. We’d ask it to build something and we’d get the most generic looking nonsense that we’d ever seen. Over time we learned with everyone else about .md files, context windows, and how to basically “onboard” an agent into a project.
The benchmark tries to measure that gap directly. How agents suck until you not only tell them about your code, but also the context around what you’re building.
I don't remember who said it first but "software teams always ship their org chart" has always stuck with me. How a team is organized has such an outsized hand it what they build.
People learning how to do things is not gatekeeping. It seems that word has lost all meaning. This person still did not actually create anything.
If you have internet access, there was never any "gatekeeping". There have been free resources to learn web development for ages. Someone not wanting to put in the effort to learn to do something is not being "gatekept", they're just choosing not to learn to do something.
Lowering the amount of effort required to accomplish some task is a good thing. Giving more people the ability to meet their own needs is a significant form of technological advancement.
This was the whole point of Visual Basic, years ago: ordinary people could build their own software. I am glad to see the same sort of thing happening once again.
They are not really accomplishing anything, though. They're not actually building anything. Meeting their needs is one thing, but we don't need to pretend like they're overcoming some insurmountable obstacle.
Beyond that, an accomplishment ceases to be one if you don't actually do anything. It's not an accomplishment if I go buy a bag of chips and eat it, even if my hunger is sated a bit.
Gatekeeping is probably the wrong word here. But like, letting someone create a script that uses regex to parse some annoying forms they have to deal with in my totally made up hypothetical is a good thing.
While there's gratification in leaning how it all works, I don't think that should be a requirement to use a computer to do a task. And the smaller that wall the better IMHO.
I get where your sentiment is coming from, but when surveying the internet and world can no longer agree. If we define gatekeeping as the requirement to extend time and effort to gain some level of competence and understanding I'm totally fine with it. "creating stuff on the Internet" has been available to pretty much everyone for a long time, and the value of their product when the creation is easy & free is also easy.
At first, we kind hated AI tools. We’d ask it to build something and we’d get the most generic looking nonsense that we’d ever seen. Over time we learned with everyone else about .md files, context windows, and how to basically “onboard” an agent into a project. The benchmark tries to measure that gap directly. How agents suck until you not only tell them about your code, but also the context around what you’re building.
reply