In other words, it is far more profitable to milk the enthusiasm of AI investors than it is to try to make money directly with the AI.
This should tell you everything you need to know about AI valuations.
2,160 karma · joined August 14, 2020
In other words, it is far more profitable to milk the enthusiasm of AI investors than it is to try to make money directly with the AI.
This should tell you everything you need to know about AI valuations.
I do not imagine that things divide so clearly along these lines.
Adding unit tests, increasing modularity, reducing duplication, and the like are all things you can tell the LLM to do explicitly. And it is happy to do it.
And you can have coding standards as part of your design criteria. Set and document these early in the project and audit the code at least enough to see they are being enforced.
I could make a full macOS clone. I could fix all the little warts that bother me and break out of the walled garden where I want to. Then I can port it to whatever hardware I like.
At this point, for software which has an extensive test suite (or a large body of interoperable software), humans are more likely to create these kinds of gaps than the LLMs are.
https://youtu.be/IyCVVkFt5-M?si=R1HROBK6zNcooXSR
I would ask why you think this, but it may be better to instead ask you to remember you said this 12, 24, 36, and 48 months from now.
The presence of instructions first added in the Pentium really does not tell you anything about performance.
Software had always been one big project with a dedicated team to build something one off, like building a hospital. Every customer commissioned their own software.
Microsoft recognized that the future was writing smaller software that was meant to be sold to many, many people for less money each.
Wrong.
What fragmentation? Find me a 64 bit RISC-V chip you want to run Python on that will not execute an RV64GC binary.
The issue is not fragmentation but ecosystem maturity. The exact same situation exists for all chips.
In most x86-64 Linux distros, CPython will be compiled to target x86-64v2. This means it will not use, for example, any of the latest vector math extensions that the x86-64v4 chip you are probably running is capable of.
The reason distros target the older profile is that not everybody has the latest hardware. There is no difference conceptually between x86-64 and RISC-V.
The difference with RISC-V is that RV64GC is more primitive than x86-64v2 and so it hurts more. And fewer people have RVA23 capable chips than have x86-64v4 chips. RVA23 is the RISC-V profile that describes essentially equivalent capabilities to x86-64v4.
But there is less RISC-V hardware in the wild in general. So, there is less legacy baggage to carry forward. RISC-V will go mainstream on a more modern profile than other chips.
Ubuntu 26.04 and newer require RVA23 support. And all new application class RISC-V processors will support RVA23. It will not be long before this is the standard RISC-V profile. It will probably happen before the x86-64 world standardizes on x86-64v4 (or even v3). So your “fragmentation” will be a bigger deal on x86-64 than on RISC-V.
But, today, projects like CPython are still using RV64GC level hardware. So, that is what they target.
Most of us do not have any desktop or server class RISC-V hardware. When we do, it will be capable of RVA23. And CPython will probably target that profile. Fragmentation has nothing to do with it.
Check back in 2 years.
You are right though that QEMU is probably faster than anything only capable of RV64GC (lacking RVA23). So QEMU would probably be a great option for the CPython team.
If all submissions were high-quality, it would probably work. But that is probably unlikely.
Maybe the project could host a searchable directory of AI contributions hosted at external GitHub repos or something.
I have been looking forward to running Haiku as a daily driver for years. And it is almost there. It may even be good enough now for many people. Modern browsers really help.
But I was amazed recently when I put Haiku on an old laptop that was already running Linux. I remember Haiku being so lean and fast and multi-tasking so smoothly. But honestly Linux felt faster on this machine. The memory usage was about the same (a real shock to me). And I missed being able to use containers. So, I ended up putting Linux back on.
This bummed me out because I have been a Haiku fan forever and it has never been better. But when I tried to use it, I could not convince myself there was any benefit (other than diversity). And the reason had nothing to do with applications or drivers.
Your clever, veiled slight, well played.
I tip my fedora.
I get why Haiku does not want to deal with it do this is not a criticism. But I have also used AI to port software and create drivers. Others may not have the ability, even with AI. Perhaps I have docs you don’t. And some of it took a reasonable amount of money and time. I am not going to dedicate that to everything I might use if it was available. And in addition to wasting duplicated money and time, we are destroying the earth recreating things others have already done.
I use Chimera Linux and I have thought about starting a repo just for the stuff that I have ported.
I am not sure what the exact answer is but I hope this is something we figure out.
The world has the technology. Canada has the resources and the manufacturing capacity. Europe has the money. It can be done.
Look at Ukraine. They are becoming a military superpower in terms of technology and experience.
The world needs leadership. And Canada is helping with that.
I do not think that reciprocal tariffs are actually the answer though. They hurt both sides and the US is not a rational enough actor to respond like you hope they will. The answer is in shifting trade away from the US and isolating their economy. It takes longer but it is what will work. And even if their behaviour does not change, their power will fade with their economy.
Well, there is some speculation regarding these chips so no sense arguing too much about it.
That said, the SpacemiT K3 is already faster than a Pi 4 and even substantially faster than a Pi 5 for multi-core and vector stuff. And that is before counting the other 8 “AI cores” it has.
These chips will be quite a bit faster than a K3, so certainly much faster than a Pi 4.