I’ve never been into his content but I really enjoyed his journey to build a better local LLM. Way more technical than I thought he was capable of. I identified with his downward spiral into madness on his journey to do something completely useless but super interesting.

Yes, but do not log out. If you do, you can’t log back in, and you can’t export. I’m paranoid so I still back up my encrypted db to cloud on a schedule.

Vaultwarden is FOSS (GPL).

It’s true re adding passwords while the server is offline, but my server runs 24x7 and it’s never down for more than a few minutes. If it goes down, I fix it. I also backup the encrypted DB regularly to cloud, so there is little risk of data loss. I am a very satisfied Vaultwarden user. Especially because it allows password sharing with my family. Everyone has an account.

Yeah it had a loose association but I loved the first person sequence.

Doom is my guilty pleasure. It’s just fun. Like Fast & Furious Tokyo Drift and Pacific Rim. I’m not watching it for the writing.

For internet access Plex is by far the easiest. You can use Jellyfin but it can be a lot more effort and can be brittle. Tailscale might be a solution but if you want to share with friends it would mean giving them access to your Tailscale network. Then you’ve got reverse proxies like Nginx Reverse Proxy. This would require buying a domain and configuring something like Cloudflare too, plus port forwarding on your router. Tailscale offers a publically accessible domain now which is similar but you cannot configure the TLD. Still, you’re opening an internet accessible port for a FOSS application and this is far less secure unless you know what you’re doing.

I think this is the last hurdle with the arr setup: discoverability. Plex has tried to jam in something, but it’s far from good. They’re never going to produce a pirate watchlist, so it would have to fall to Jellyfin. What people are seeking is the Netflix experience of “curated” content, spoonfed, and instantly watchable.

For the record I do the same as you. I think the intentionality is a healthy barrier to mindless browsing and consumption, but once people are hooked, it’s hard to wean them off.

There is no evidence that lead poisoning, for example, can be inherited. There is some research that certain epigenetic changes can be inherited. The context is sometimes war and famine. But that is 2-4 degrees of abstraction away from your question, and we are many decades away from any kind of causative conclusion to that.

If I were to steelman your position, I think there are simpler changes to point to as potential causes here. Social media. Screen time. Time spent outside and exercising. Obesity. Time spent socialising. The new “permissive” method of parenting and teaching.

You’re going to have to elaborate on what a genuine disability is there chief. Let me help you out:

The UK Equality Act defines a person disabled if they have a physical or mental impairment, and the impairment has a substantial and long-term adverse effect on their ability to carry out normal day-to-day activities.

I believe all of those things you listed were much worse in the 70s (with the potential exception of microplastics) when disability rates were much lower. There is no proof that microplastics are causing autism and ADHD (and a thousand other disabilities). What has changed is diagnostic criteria. In the UK they have become much looser.

Why is there never any nuance in these discussions? We can both believe that under-diagnosis occurs, and that over-diagnosis occurs. 20% of all pupils in the UK are now classified as so disabled that they require specialised assistance. “SEND” assistance for this can range from free taxi services to and from school (which recently reached £1.2 billion), to support payments, to special assistants in school. The number of ECHP students (those with the highest needs) increasing by 71%, from 253,679 in 2018 to 434,354 in 2024. SEND spending is out of control.

So what happened, exactly? The average child disability rate in Europe is 4.6%. How did the UK end up with 20%? Did the UK suffer a catastrophic nuclear event? A war? Famine? None of the above. It is clear that categorisation has become EXTREMELY loose over time on average. This does not mean that there are not children who are struggling to get diagnosed with ADHD. However ADHD and autism are a spectrum disorder. It is not binary. The UK has drawn the line far closer to the normal side of the spectrum than any other nation on Earth. If costs continue to rise at this rate, it risks destabilising the entire health system. Public sentiment will shift, and we risk undermining children getting any diagnosis at all.

IMHO, this requires at least two tactics at the same time. 1) Invest sufficiently into diagnosis resources. Stringing parents and children along for years while they wait in the system can make the issue much worse than it needs to be. 2) Draw the diagnostic line closer to where the rest of Europe does it. This will mean far fewer children are diagnosed with disabilities, but those who genuinely have a disability are treated much faster and actually receive the resources they need.

New York spends $29,873 per student (the highest in the country). California spends $16,739 per student.

Utah spends $9,496 per student and has higher NAEP 2022 Grade 4 Reading average (221 vs 214 for NY), and NAEP 2022 Grade 8 Math average (282 vs 274 for NY).

Florida spends $11,681 per student and has higher NAEP 2022 Grade 4 Reading (225 vs 214 in California), and higher NAEP 2022 Grade 8 Math: (271 vs 270 in California).

The issue isn’t funding. It’s also not ideology, since there are clear examples of red states with better educational outcomes than blue states. All while spending much less per capita.

Okay that is weird. My comments were more real life experiences. Then include me in the mystery. Why would people oppose fibre?

I would be surprised if that were true but I haven’t seen any data either way on that.

Quite right, and this is the problem with correlational studies like that linked: they’re observational. We can’t do controlled studies on humans, so we try very hard to control for confounds.

“I refuse to look at text on the internet.”

“I don’t like the messenger so I’m going to reject their proposal for a healthy diet.”

Cool story, bro. I hope you enjoy your cheeseburgers and diabetes. It suggests a healthy diet of protein, vegetables, and healthy fats. I presume that would be an immense improvement over yours.

Do you chug it from the can or add some spices?

This is where the data meets an uncomfortable impasse. Hispanics in the U.S. live a lot longer than whites. It’s not because of socio-economics. Hispanics eat a lot of meat, too. It is theorised that their legume heavy diet might be responsible. Meaning that, despite eating a lot of carbs, they still live longer. I suspect longevity effects are primarily related to the fibre. This might be true in the study linked. Perhaps it isn’t the low carb/low fat part which gave heart benefits, but the higher fibre intake.

As for the mechanisms:

1. Fibre binds bile acids, which increases bile acid secretion. The liver uses LDL to make more bile acid, reducing LDL.

2. Fibre acts as a glucose release modulator. Glucose spikes are responsible for a range of immediate and systemic issues. Reducing overall glucose consumption and especially spikes results in lower risk of type 2 diabetes, metabolic syndrome, and atherosclerosis progression.

3. Lower blood pressure (we’re not really sure why yet).

4. Fermentable fibres are metabolized by gut bacteria into short-chain fatty acids. These reduce systemic inflammation, improve gut barrier integrity, improve insulin sensitivity, and regulate immune function. The inflammation one in particular could explain a great deal of the health benefits.

5. The cancer risk specifically decreases because of dilution and faster transit of carcinogens, SCFA production (especially butyrate, which has anti-tumor properties), and reduced insulin and IGF-1 signalling.

6. Fibre keeps you full for longer, reducing the risk of obesity.

But convincing people to eat fiber gets a weird amount of opposition

I haven’t seen much. It’s usually when it comes in the form of vegan evangelism and moralising. “Eat a balanced diet which include meat and veggies” is about as milquetoast a message as you can get with very little opposition. In fact, the new U.S. food pyramid is exactly that. Reception has been very good. It’s not perfect, but it’s a hell of a lot better than the old one.

I’m not presuming anything. They explained their intent. It was the EPA which ignored Congress’ clearly articulated intent.

you’re not only presuming congress’s intent

No, again, read my comment. Congress specifically chose not to regulate CO2 in the Clean Air Act Amendments of 1990. It was discussed and rejected. It is a matter of public record. No assumptions need to be made. We know exactly what was intended to be in the legislation.