kookamamie

Exactly. There seems to be this fantasy in which you can somehow string different kinds of agents together, one designing and one reviewing, and that finally producing something superior as output - I just don't buy that.

Sounds like heuristics added on top of statistics, which is trying to remedy some root problem with another hack.

Sure, but considering the things such a workstation might be used, I'd say the gap is closing between the segments.

The memory channel difference is a big one, and obviously that is not stressed fully by a rendering test like Cinebench as there's very little memory traffic involved.

The 9955WX hardly scores higher than my 9950x3D in Cinebench - yet, it's 2-3x in price, not counting in the pricier mobo.

Yeah, you're right. Depends on what's in the AVI, of course. If the input is already something that can be housed inside mp4, i.e. standard codecs, the packets can survive by just copying them.

> --- Generated ffmpeg Command --- > ffmpeg -i my_video.avi -an -c:v libx264 my_video.mp4

The example itself shows a naive conversion, ending up transcoding to default h.264 params. This should have been -c:v copy for copying the input packets, as-is.

> On my M2 MacBook, the renderer process is now using 6% CPU

Looking at the animation, I'd wager it's still way too much CPU use for what it is.

> 37 different products

I guess we have different understandings of what a product is.

The animation worked for me, I'm red-green colorblind.

As a bonus the LLM can spew out endless amounts of bullshit.

Applogies! You're exactly right, here's how this spans out…

I was there, too. Not at Nokia, but in the ecosystem of these companies.

People wanted iPhone over Nokia, not due to its specs but due to its usability and presentation.

Let's be honest, both Symbian and Maemo/Meego were abject messes in both of these categories.

Quite nice. Although, it seems to be making some pretty interesting assumptions. For example, most operations seem to assume ARGB inputs are premultiplied-alpha and will call PLUTOFILTER_UNPREMULTIPLY_PIXEL.

One can dream.

You shouldn't worry about LinkedIn, the cancer of the internet.

> Thunderbird

Ahh, does it still have the bug that may accidentally delete/corrupt all your emails?

Looks like a bubble, pops like a bubble.

> The puzzle is why

The breakthrough isn't in their datasets.

Renoise is very cool - I've used trackers since the early 90s and Renoise captures that experience just right.

No, it is the way. Edit: no one has time for inventing unique names for include guards.

Agreed about C's aliasing rules. Fortran had a better set of defaults.

I found it unnecessary to highlight their age.

Yes, makes sense and I don't mind at all - was part of the same scene.

> Information to warez groups

Oh wow!

You're now officially working for the machine, congrats.

I don't think the native C++, even when bundled with OMP, goes far enough.

In my experience, ISPC and Google's Highway project lead to better results in practice - this mostly due to their dynamic dispatching features.

Agreed, picking Rust for the project likely has very little to do with "memory safety".

Yes, exactly. See Eigen as an example.

> Bump-maps

The terminology seems rather old.

> auto & rsquared = expression.Mul(expression.GetP1(), expression.GetP1());

This is C++, no? Why not use operator overloading for the project?

I hope this will happen, too. I think it might as soon as investors realize the LLMs will not become the AGI they were sold as an idea.