> The discovery that chronic inflammation can provide the impetus for cancers to develop is forcing clinicians to rethink their approach to the disease’s prevention.
Alternative health has been saying this for decades. Ketogenic diets + medicinal plants/mushrooms can do a lot, even after the fact.
There's an awful lot of stuff that works, that nobody has run a large enough controlled study to prove it works. The organizations which fund medical research have specific priorities that exclude an awful lot. And a lot of things are just inherently difficult to objectively measure or control. There's no blood test for chronic muscle tension, for example.
So unfortunately, by restricting yourself to things that have been proven to work, you are possibly eliminating a lot of things that work.
But of course, trying to figure out, on your own, which stuff actually does work despite not being proven, is a long hard frustrating slog that tends to involve a lot of personal trial and error. Exactly what GP said:
> The difficult part is figuring out what's true and what's quackery.
Alternative medicine is simply any therapy that is not included in the established currently-accepted set of treatment options.
This varies by culture, time, and sometimes by individual.
Most alternatives are not better than the currently-known best. This is true today, we think, but it is definitely not true historically. (So how special is our current era?)
But when the currently-known best doesn't work well for everyone, or has deleterious side effects, any continued research will include alternatives.
I understand the fatigue embedded in your quote. It's a reasonable stance for those of us with ordinary concerns and who are far downstream from the research (including and especially retail practitioners).
But it is too broadly dismissive for real scientists and people who maintain a curiosity about the world.
It's not hard to say all kinds things if you don't do it with scientific rigor. Unfortunately, it's very difficult to know which of the things are true.
Inflammation is typically an immune response. Chronic inflammation is a standard flag that is measured a used by doctors in diagnoses for many years, and by oncologists in particular when treating cancer.
There’s a lot of work right now into immunology and cancer, and they are discovering specific correlations as that progresses. This has nothing to do with mushroom tea, although that probably helps with acute inflammatory issues.
As far as I understand, Keto helps manage the symptoms of diabetes, but will not prevent it, and (without a source on hand), might actually increase your chances of developing it relative to a healthy diet that includes (mostly unrefined) carbs.
Nix (and nix-darwin for persistent config) already does the job for non-GUI apps better than anything else could (its catalog is unmatched), and homebrew (also available in nix-darwin) handles GUI apps.
I doubt nix is ever going to enjoy popularity on the scale of Homebrew due to its differing model. People are too used to doing a simple “x install y” without any backing configuration or flakes or anything like that.
Personally speaking I might learn it at some point but for now nix's friction/overhead isn’t proportional to the occaisional inconvenience with traditional package managers.
I think Nix has a marketing issue as it goes so deep. But if you're going to consider it only as an alternative to homebrew, you don't have to know anything aside from different CLI invocations: `nix-shell -iA <package>` to install, and maybe a step up with `nix-shell -p <package> --run <cmd>` to run without installing. You don't have to use nix-darwin or managing configs and the Nix language at all.
Yeah, I'm just addressing you specifically, you could switch easily and Nix has a much larger package selection (outside of GUI apps on macOS). And it immediately can do more via nix-shell -p, which can be also be used as shebang.
But yeah, Nix is much more than a homebrew replacement and that has its downsides.
I'm curious of what the packages are because I only have ~10 packages that I use from homebrew, either very Mac-specific or that have better packaging (ffmpeg, mpv)
They are all projects that distribute Go binaries on GitHub and have official homebrew taps
The 2/3 that are on nix are both out of date and created by some outside party.
I do not see a world where I provide nix stuff for my projects. There are already too many packages managers and I'm not spending my unpaid time to support all of them. GitHub and brew are my limit, def not learning nix just for this
If the complaint is that other systems are less up-to-date than brew is, then the number of things installed is still the issue - whether they're dependencies or not.
Unless, of course, the other system manages dependencies differently, say it compiles in static versions of required libraries at build-time, or divides them up into fewer dependent packages.
I tried nix a few times in the past and it has too many issues on Mac (never got it working), also the learning curve is too steep, maybe I'll give it a try again.
... but it "does the job better than anything else could".
---
This is the issue I have with the whole nix discourse: it's hailed as this great magical tool. But when pressed for details it's always this: weird commands, configs, working or non-working wrappers etc.
This is just from my experience, I'm not part of any kind of Nix cult or anything, not that I'm aware of.
I don't know what you're talking about re: third party wrappers, this is just an option for a slightly easier and perhaps more robust install. And homebrew isn't free of issues at all.
So a "better than anything" still needs "a better more robust install" and you still use homebrew for "packages that are packaged better like ... the extremely popular and often installed ffmpeg"
;)
> And homebrew isn't free of issues at all.
No one advertises brew as "the best", "better than everything else" etc.
BTW, looked at Determinate Nix. It's a custom distribution and a custom daemon, and its own fork of documentation
I don't use Determinate Nix but I've heard it recommended. As I've said I've had 1 or 2 issues with my Nix install on Mac over several years.
Re: ffmpeg, it's just something do to with default codecs or something I don't remember, I'm not even sure I really needed it but I saw it on my packages list and since I use nix-darwin it doesn't matter that much where the package comes from. I have hundreds of packages from nixpkgs, many shared with Linux, and quite a few Mac specific ones.
Again, I'm not part of a Nix cult, I just genuinely have had an overall great experience with Nix and I find the criticism overblown. I guess since a subset of users are very enthusiastic and vocal about it it attracts this.
Before getting a new MacBook, I was using MacPorts since Homebrew had several problems, including what the GP mentioned. For over a year now, I have used home-manager with Nix. Nixpkgs has the "important" subset of GUI programs that I use. Major exception to this is OrbStack, but there is a homebrew-to-nix converter that I use to handle that.
I briefly experimented with nix-darwin, but concluded that it was overkill for my needs. Home Manager also has the benefit of not being Mac-specific (so plenty of questions and examples exist online).
Most importantly, I no longer have to mentally prepare myself when updating Mac OS. If I recall correctly, both MacPorts and Homebrew require the user to re-install all installed packages after a major version update, but this problem doesn't really exist on Nix. It's also nice being able to store flakes in a Git repository and effortlessly rollback changes or replicate a whole development environment across several machines.
I'm interested in how you use that homebrew-to-nix converter, if you have links to configs or can describe it briefly!
I agree that nix-darwin is a bit overkill, I mainly use it for the homebrew integration, but I think so is home-manager, I don't keep user app configs in Nix since they are never exhaustive and more effort than it's worth.
Coming from nixos, I didn't want to use brew but I found nix-darwin to be poorly documented and supported.
Several upgrades wrecked havoc on my system, a lot of things didn't work as expected, I had to package a lot of stuff; eventually I just bit the bullet and started using homebrew
I sort of get what you're saying and I've had a few sporadic issues myself, once or twice I had to reinstall (although might have been my mishandling), but overall I can't see a better solution than this. Plus I can share the configs with Linux.
You don't strictly need it, it just makes it a tiny bit more convenient since you can set it up to override DNS on any connected device, and Tailscale sets up a private VPN mesh between your devices I've come to get take for granted - a tangential feature that goes well with centrally managed DNS.
There's a lot more to Tailscale but for a basic setup you just install the client on all your devices, and set DNS to the NextDNS endpoint. Any device on your network will automatically pick it up.
People will develop an eye for how AI-generated looks and that will make human creativity stand out even more. I'm expecting more creativity and less cookie-cutter content, I think AI generated content is actually the end of it.
>People will develop an eye for how AI-generated looks
People will think they have an eye for AI-generated content, and miss all the AI that doesn't register. If anything it would benefit the whole industry to keep some stuff looking "AI" so people build a false model of what "AI" looks like.
This is like the ChatGPT image gen of last year, which purposely put a distinct style on generated images (that shiny plasticy look). Then everyone had an "eye for AI" after seeing all those. But in the meantime, purpose made image generators without the injected prompts were creating indistinguishable images.
It is almost certain that every single person here has laid eyes on an image already, probably in an ad, that didn't set off any triggers.
Given that the goal of generative AI is to generate content that is virtually indistinguishable from expert creative people, I think it's one of these scenarios:
1. If the goal is achieved, which is highly unlikely, then we get very very close to AGI and all bets are off.
2. If the goal is not achieved and we stay in this uncanny valley territory (not at the bottom of it but not being able to climb out either), then eventually in a few years' time we should see a return to many fragmented almost indie-like platforms offering bespoke human-made content. The only way to hope to achieve the acceptable quality will be to favor it instead of scale as the content will have to be somehow verified by actual human beings.
> If the goal is achieved, which is highly unlikely, then we get very very close to AGI and all bets are off.
Question on two fronts:
1. Why do you think, considering the current rate of progress think it is very unlikely that LLM output becomes indistinguishable from expert creatives? Especially considering a lot of tells people claim to see are easily alleviated by prompting.
2. Why do you think a model whose output reaches that goal would rise in any way to what we’d consider AGI?
Personally, I feel the opposite. The output is likely to reach that level in the coming years, yet AGI is still far away from being reached once that has happened.
1. The progress is there but it's been slowing down yet the downsides have largely remained.
1.1. With the LLMs, while thanks to the larger context window (mostly achieved via hardware, not software), the models can keep track of the longer conversations better, the hallucinations are as bad as ever; I use them eagerly yet I haven't felt any significant improvements to the outputs in a long time. Anecdotally, a couple days ago I decided to try my luck and vibe-code a primitive messaging library and it led me in the wrong path even though I was challenging it along the way; it was so convincing that I wouldn't have noticed hadn't my colleague told me there was a better way. Granted, the colleague is extremely smart, but LLM should have told me what was the right approach because I was specifically questioning it.
1.2. The image generation has also barely improved. The biggest improvement during the past year has been with 4o, which can be largely attributed to move from diffusion to autoregression but it's far from perfect and still suffers from hallucinations even more than LLMs.
1.3. I don't think video models are even worth discussing because you just can't get a decent video if you can't get a decent still in the first place.
2. That's speculation, of course. Let me explain my thought process. A truly expert level AI should be able to avoid mistakes and create novel writings or research just by the human asking it to do it. In order to validate the research, it can also invent the experiments that need to be done by humans. But if it can do all this, then it could/should find the way to build a better AI, which after an iteration or two should lead to AGI. So, it's basically a genius that, upon human request, can break itself out of the confines.
People already know what the ads are and what is content, but yet the advertisers keep on paying for ads on videos so they must be working.
It feels to me that the SOTA video models today are pretty damn good already, let alone in another 12 months when SOTA will no doubt have moved on significantly.
Alternative health has been saying this for decades. Ketogenic diets + medicinal plants/mushrooms can do a lot, even after the fact.
reply