I believe that it's more nuanced than that (this is from the context of German regulations):
In the old days, reflectors and diffractors were rather crude and light distribution wasn't even at all. Regulation was setting a light intensity limit for the arc above the cutoff, assuming that the brightest spots would be just that, unintended brightness outliers. And when those would comply with the limit, the typical brightness hitting the eyes of oncoming traffic would be much lower.
Now reflector design is SO much better manufacturers striving to make their buyers happy can make lights pushing out photons exactly at that old regulatory limit over the entire cutoff area, and with a precise jump where the cuttoff stops. That's awesome for the person behind the wheel ("best lights I ever had!"), technically within the old limits and a terribly blinding for everybody else.
Looks very much like the only chance of that ever happening now is if someone established a separate league that only allows naturally conceived horses.
I think that what they mean is that instead of ten perfectly orthogonal "unix philosophy" tools (skills) for the agent to compose when solving a problem, each with an API surface (description text) the size of Texas, you'd want to can each composition in a shell script (or a bespoke rust binary, if you enjoy watching your bot perform some heavy lifting) that only solves one problem but solves it so focused that the accompanying skill description barely consumes more context than the tool's self descriptive name.
I still didn't follow, you mean to pipe things between tool calls? Like if you want to query something and then update another without the intermediate getting brought in context?
Instead of requiring each session to understand the n tools used to solve a particular problem, you bundle up the solution in a conventional script (that's what I meant by "can", as in canning) that the agent can use with very little documentation in the context. When the model is smart enough to figure out the composition of underlying tools during regular execution, it will also be able to do the canning up as a script and write the lightweight documentation that turns the script into a skill. Subsequent use will only require that lightweight documentation in context.
Won't you just end up with hundred of very specific scripts that can only do a very narrow thing? And now they'll all have their description and name in context.
Is that so surprising? I thought that was a given. And as soon as remote resources are involved, the old "it's in a docker" peace of mind does not apply.
Read about it back then in what might best be described as a kid's annual hardcover magazine and loved it so much (a few years later they had a long article about the Porsche 959, it's almost a tie for me). "Das Neue Universum": in those years they had an awesome mix of technology, culture, adventure and science - some parts much further from "ELIF" than one might expect.
A part of that young me still seems to live on being mighty disappointed that I'm not living in that future!
Do you find it hard to max out, or do you find it hard to productively max out?
It's like paying drivers per gallon of fuel consumed and then acting all surprised that you see them revving their engine while waiting at a red light.
In the old days, reflectors and diffractors were rather crude and light distribution wasn't even at all. Regulation was setting a light intensity limit for the arc above the cutoff, assuming that the brightest spots would be just that, unintended brightness outliers. And when those would comply with the limit, the typical brightness hitting the eyes of oncoming traffic would be much lower.
Now reflector design is SO much better manufacturers striving to make their buyers happy can make lights pushing out photons exactly at that old regulatory limit over the entire cutoff area, and with a precise jump where the cuttoff stops. That's awesome for the person behind the wheel ("best lights I ever had!"), technically within the old limits and a terribly blinding for everybody else.
reply