I really think the authors in FT are not considering:
- The innate value of the models come from not training chips nor hosting, but the actual quality of thinking traces, inference costs and usage. Anthropic shines here, but not for long.
- The variety of skills and evals on new tiny open models proves Claude is expensive and not extremely better than these models today, and math benchmarks alone do not cut it.
- Enterprises are realizing that they can just buy some regular Macs and do their own inference far cheaper than invest in Claude Subscriptions, and Claude given tools are losing their edge to harnesses such as Pi and Opencode.
- Nobody needs Hacking Tools unless they are red-teaming their own code. The niche of people wanting this could write their own tools and workflows to test the effectiveness of many LLMs on these workflows and see how far they can push it.
- The amount of collusions, hypes and fear mongering do not sit well with many people (including myself) - stick to the science and make usage safe, like everything else we do with software in production.
- The IPOs often get priced down with market discovery and anyone who's buying it ONLY is interested in the long run, not inflated valuations not hype not any of these lies used to sell this stock. I am bearish about the IPO at this rate.
- For me, reasons for SpaceX money investments are NOT due to AI *That part is a true waste of money* but good enough due to manufacturing, supply chains and actually building useful Satellite infrastructure; I still am bearish about them launching NVL72 nodes in space, not because of the space part - I just don't trust those chips enough for it to sustain that long.
LLMs were a bad idea. MCPs were the band-aid to get them closer to the real world and do something for recursive self improvement. Then LLMs became better suited for everyday use.
I'm tired of these kind of posts. This author is on my black pill list.
It's great they do C23. I'd love to get three-four projects worth of things merged together for a good reason:
- TinyCC, SDCC, Musl and Cosmopolitan Libc
Reason is that I require a quick fast portable compiler with a decent runtime that can also be embedded into a larger runtime, like the way Bun does - while keeping runtime small and portable.
SDCC does not need the same amount of bloat - having decent small C runtime is a job in itself. Fixing it once could allow a better compiler that isn't quite as bloated as LLVM nor as small like TinyCC.
Many developers will likely be developing at Android 16 APIs only. Will rely on Aptoide+AOSP exclusively. This is likely the timeline we'll see a decline in Play Store and Pixel usage.
We all know these aren't purely technical reasons, it's to break the monopoly of one Internet Company's decoder format portfolio. You can't let one Fruit Company's format be the thing either, so it's basically a L for engineering but a W for pleasing everyone TBF.
AVIF is royalty-free, with open-source implementations not developed by Google. Google's implementation is libaom; as you can see, SVT-AV1 beats it. SVT-AV1 is developed by a number of companies – historically Meta, Netflix, Intel, independent contractors, and others. I worked on SVT-AV1 myself. I don't see how this is a bad thing for everyone, even if Google drove standardization of AV1?
There were something like 40 companies involved in AV1 standardization. Quite a few technical contributions did not come from Google, and quite a few of Google's proposals were substantially modified by feedback from others before they were included (and not all of them were eventually included).
Source: I lead Mozilla's technical contributions to AV1.
It saddens me to see discourse on the internet about AV1's ancestry being focused on Google's VPx line of contributions, while diminishing those coming from Daala (entropy coder...) and Thor (QMs...), as well as inter-company collaborations (CDEF).
A redeeming outcome is that with AVIF's new image tuning modes (both in libaom and SVT-AV1), Gianni and I managed to utilize as many AV1 coding tools as possible, including QMs that sorely needed a well-deserved spotlight.
Also, thanks for paving the way to the current state of multimedia compression! I'm a longtime fan of Xiph(ophorus) since the 1.0 beta/RC Vorbis days. We used your image sets a lot during our testing.
You're welcome. That is what those image sets were for.
I was really happy Thomas Davies added quantization matrices to the spec, because if he hadn't, I would have had to. Fortunately, he did a good job. I am glad someone is finally using them. Don't sleep on per-segment quantizer deltas, either.
Me and a bunch of friends (mainly Clybius and Yiss/Uranite), after I was constantly saying daala was a great low complexity visual metric that should be added back to AV1 encoders, we decided to implement it into yet another svt-av1 encoder fork:
https://github.com/Uranite/svt-av1-tritium#from-svt-av1-hdr-...
It has been showing great performance overall and is quite nice.
It can work on CDEF, like it did in aomenc-av1 back in 2018/early 2019. It does so very well.
We've also implemented deeper controls, but that is still experimental.
Thank you so much for your work on daala stuff, as it's been one of my many inspirations in the field.
HEIF isn't an Apple owned format, most modern cameras have it as an option alongside JPEG now. The situation is JPEG is woefully outdated, lacking in HDR support and dated compression. HEIF is technically much better but patent encumbered.
JPEG-XL offers all the benefits of HEIF and more, while not being patent encumbered.
I'm also thinking the OP meant HEIC, but I know that it is not proprietary (though as noted patented), and wanted to make sure, as a lot of people of all sorts of assumptions/misconceptions about Apple.
HEIF/HEIC is based on H.265, so not Apple-only as Android supports it. (Recent Sony, Canon, Nikon, etc cameras all support the format as well.)
Don't buy this argument. It's like saying, we lords who hold this capital will build all this economically destructive stuff anyway, and wait for the world to not react to our stupid ways of enriching yourselves.
People are not going to slow down because this was brought to them on less than endearing terms. They don't see any of these stated noble intentions.
Claude was already used to cause economic, political and social destruction. As Anthropic is seen doing it, others aren't going to just sit down, read the blog and say, oh, I'll stop developing models because Dario, you touched my heart with your true words.
I also don’t buy the arguments. The arguments about helping humanity cure diseases as if those are some of the leading Causes of suffering. The bar is so low that we don’t event need to cure diseases, we just need to lift the entire floor with basic solutions like: housing for everyone, continued education on healthy eating and just general continued education on like living. Life coaches so many very basic things that don’t require a novel solution or a data center.
Solving a disease like cancer as justification for everything else seems like not a great trade off. I don’t say that to dismiss those who have lost people to cancer. But the technology just is being utilized in so many harmful ways and the things it enables: job loss, surveillance, misinformation, just feel like problems that aren’t worth it. Have we even heard of a solved diseases? If anything haven’t we heard that ai will make even worse bio warfare?
Yeah, it's insane people are using the argument that we'll pour trillions on moonshot cancer research when pouring trillions into universal healthcare would probably save countless more lives.
It's this kind of argument where the mask sort of slips because it always has to be an argument that also has the benefit of enriching himself and his friends.
- The innate value of the models come from not training chips nor hosting, but the actual quality of thinking traces, inference costs and usage. Anthropic shines here, but not for long.
- The variety of skills and evals on new tiny open models proves Claude is expensive and not extremely better than these models today, and math benchmarks alone do not cut it.
- Enterprises are realizing that they can just buy some regular Macs and do their own inference far cheaper than invest in Claude Subscriptions, and Claude given tools are losing their edge to harnesses such as Pi and Opencode.
- Nobody needs Hacking Tools unless they are red-teaming their own code. The niche of people wanting this could write their own tools and workflows to test the effectiveness of many LLMs on these workflows and see how far they can push it.
- The amount of collusions, hypes and fear mongering do not sit well with many people (including myself) - stick to the science and make usage safe, like everything else we do with software in production.
- The IPOs often get priced down with market discovery and anyone who's buying it ONLY is interested in the long run, not inflated valuations not hype not any of these lies used to sell this stock. I am bearish about the IPO at this rate.
- For me, reasons for SpaceX money investments are NOT due to AI *That part is a true waste of money* but good enough due to manufacturing, supply chains and actually building useful Satellite infrastructure; I still am bearish about them launching NVL72 nodes in space, not because of the space part - I just don't trust those chips enough for it to sustain that long.
reply