You can't blind a trial where your diet is a 3:1 ratio of fats to non-fats. There just no sugar pill for that. There's no self-selection here either- just eligibility evaluation followed by informed consent (because people are not mice). A placebo control arm would be immoral due to the pace of this disease leaving no option for crossover.

  As for sample size, the sheer size of the effect blows away your objection. Ask yourself this: "There are 18 glioblastoma patients. Seven of them have stayed alive 33, 36, 36, 84, 43 and 44 months after diagnosis. Calculate the odds of this occuring by chance". Answer: 0.1 to 1.2%, making it extremely unlikely to happen by chance. Even if you add 2 more people to N (assuming they were ineligible due to comorbidities), the answer stays the same: VERY UNLIKELY. This is even before you analyse the compliant subgroup.

Small N is not a problem when the "Effect Size" is this large. In statistics, if you are looking for a tiny 2% improvement in survival, you need thousands of patients to prove it isn't an accident. But if your effect size is massive like this one is, you only need a small group. A 3-year survival rate jumping from 8% to 58% (compliant subgroup), or even from 8% to 25% (naive analysis of overall group) is a massive signal. It is like testing a parachute: you donâ(TM)t need a randomized trial of 10000 people jumping out of planes with and without parachutes to prove they work; the effect size of the parachute is immediately obvious in a small sample jumping with parachutes (It's also the moral trial to do).

I linked to one below

Yes, my blog is terribly written. :-D

But the rest of what you says is utterly wrong.

Have a look at the paper I pointed earlier in this thread on patients kept alive using his protocol.

The basis of Dr. Seyfried's work is that the 'nuclear DNA mutations cause cancer' theory is wrong. It's the mitochondria. This is shown by research where nucleus from cancer cells transplanted into normal cells resulted in a normal cell. And when mitochondria aretransplanted from a cancer cell to a normal cells, that gave rise to a cancer cell.

You can't fault a man for who he happens to talk to, otherwise you'd be condemning Jesus too. I don't know what he told Mercola; maybe "here's a smarter way to think about this"? :)

Have a look at the paper I pointed earlier in this thread on patients kept alive using his protocol.

The basis of Dr. Seyfried's work is that the 'nuclear DNA mutations cause cancer' theory is wrong. It's the mitochondria. This is shown by research where nucleus from cancer cells transplanted into normal cells resulted in a normal cell. And when mitochondria aretransplanted from a cancer cell to a normal cells, that gave rise to a cancer cell.

It sounds like you're still recoding it. I believe the design goal you are referring to is *absolutely no change* in the image. In fact, it goes futher: I believe there's a path to convert a JPEG to a JPEG XL that is typically much smaller such that you can undo the conversion from the JXL and get back the exact bit-identical JPEG file you started from.

Here is one:

https://www.frontiersin.org/jo...

Kiryttopoulos and colleagues (2024), "Successful application of dietary ketogenic metabolic therapy in patients with glioblastoma: a clinical study," Frontiers in Nutrition, vol. 11, article 1489812. Seyfried is a contributing author.

Thanks for the tip!

Yes, that's because I put some useful cancer info there. Don't let this upset you - I earn no dime from that blog.

One of the most important info there is on Dr Seyfried work. His approach has kept patients alive - one pancreatic cancer patient alive for 6 years and a glioblastoma patient alive for a couple of decades. His approach relies on continuous glucose and ketone measurement. I've calculated my GKI using a glucose meter, and it can get complicated pricking yourself twice, running two test strips, doing the ratio, etc.

Here's a good overview on Dr. Seyfried's work:
https://pubmed.ncbi.nlm.nih.go...

I am promoting Dr. Seyfried. I am not Dr. Seyfried.

Dr Thomas Seyfried's work on dietary approaches to treating cancer uses Glucose-Ketone Index (GKI) -- a simple way to track how effectively you're shifting away from glucose metabolism toward ketone metabolism. The cancer patient makes dietary decisions to drive their GKI numerical value in the low single digits.

GKI Calculation:
GKI = Blood glucose à Blood ketones

Target Ranges:
General cancer management: GKI 2-3
Active treatment: GKI 1-2
Advanced/aggressive cancers: GKI below 1 (temporarily, with medical supervision)

Normally, you test fasting glucose and ketones upon waking. But continuous measurement provides a lot more insight.

I wrote a blog about GKI and other 'self care' approaches here:
https://enconsed.blogspot.com/...

The screen that matters is the windscreen. Nothing should compete with it

Correction to my typos above: "The plates are both: on public display, and meant to provide anonymity (*) to vehicle owners, drivers and passengers"

(*) Not including plates such as "LEON-72"

It's not the ALPR. It's the registration lookup that's the problem. ALPR - the core tech - is genuinely useful. Consider this thought experiment: normal people (you, me and our neighbourhood posse) implement our own ALPR network. We run it on poles in frontyards, on old mobiles duct-taped to windows, on personal phones in its cradle as we drive around. We run torrents that sync ALPR data with trusted nodes. Someone vibe-codes an app that correlates ALPR data with public crime report: It flag plates possibly associated with crime, such as the casing of an area before burglaries. Another app detects bad driving (yes, where someone's "bad" could be someone else's "snazz").
So far, all is functioning as it should. Capturing ALPR data privately is just photography in a public area with no privacy violation. The plates are both and provide anonymity to vehicle owners, divers and passenger. The private ALPR network could help report concerns about specific plates to the police with no idea who owns the plates.

What breaks privacy is running a plate/car registration. Now you know the owner, where they live, whether they have insurance. Doing *that* should require lawful exigency or a warrant.

Don't trust the LLMs we've cooked up currently:
LLMs: Untrustworthy Computing
 

I agree. For an entirely different reason:
LLMs: Untrustworthy Computing