News for hackers. Stuff that matters.

VernorVintage

Re:Llama.cpp 30B runs with only 6GB of RAM now on Saturday April 01, 2023 @03:55PM (#35403525)
It's a pretty big concern if you had to spend a billion on training, but 6 months later the open source community is able to replicate your training for <100K because you were too cheap to hire an extra 100 optimization experts

That'd be a 10,000 fold depreciation of an asset due to a preventable oversight. Ouchies.

Re:Moving from Rust to C++ on Saturday April 01, 2023 @03:37PM (#35403357)
What was one of the benefits of C++ that was discussed? It seems like everything mentioned also comes with a terrible faustian bargain. For instance, C++ build systems are as plentiful as they are byzantine and torturous. Some new language features are heaven, but when your compiler hasn't implemented them yet, suddenly you're in hell. And the fact that bigots tend to feel freer in C++ communities than in rust communities certainly feels like a deal with the devil.
Re:Moving from Rust to C++ on Saturday April 01, 2023 @03:28PM (#35403283)
The part where they said they didn't need inclusivity in computer science made me wanna cry because I didn't realize it was april fools and there are so many people who make that argument unironically..

also the part where they promoted make/cmake :((((

or claimed they can do unsafe code just fine...

all these are real world positions that i've heard defended by old C++ barnacles v_v

Re:Moving from Rust to C++ on Saturday April 01, 2023 @03:23PM (#35403236)
No, bad. Bad. Do not propagandize for the office, even if you miss it. squirts u with windex BAD CAT!
Okay, wow so HR 1.8 literally does mean that fasters are nearly twice as likely to die over a given period as non-faster. But the death rate for fasters and nonfasters can still be pretty low, for example 2% a year vs 1% a year.

Since the risk is low already, an individual might be indifferent to a 1% or 2% death rate - from an individual perspective the two are even statistically indistinguishable (they can't exactly die more than once!)

Okay, I'm a little less freaked out by the seeming /doubling/ of death rate, however, it's still 2x increase in death rate. Still chilling!

Now I suppose that if we assume the death rate is constant for everyone at all times - or use a more realistic actuarial curve - then we can even transform the HR factor into an average effect on lifespan for fasters vs nonfaster.

Then we could say "ah, nonfasters will die on average 4 years sooner than nonfasters". I think that would help contextualize the numbers even better than relative death rates (which are hard for individuals to reason about).

Are you familiar with how we might convert death HR into an expected lifespan delta for the average person? Would that analysis actually hold any meaningful truth?

The article mentions feeding periods of <4.5 hours, which is a well known intermittent fasting protocol, which I have already defined. You have willfully ignored that, suggesting you might not respect my time or the truth.

I caution you that you are using a nonstandard definition of fasting that does not include all the currently recognized forms of intermittent fasting. Nor have you presented a definition of intermittent fasting - it's likely your definition is too narrow to cover the current practices.

I refer you once more to the wikiepdia article on IF (which leads . Again, the relevant term is "Daily Restricted Feeding". That information is here:

https://en.wikipedia.org/wiki/Intermittent_fasting

It's entirely fair to regard the current research as having to do with Daily Restricted Feeding, which is a form of IF.

And yes, the article does not mention snacking. I was simply claiming that some IF protocls even allow light snacks, to help persuade you that IF protocols are more diverse that you think.

The goal of these protocols is to reduce things like insulin resistance, which happens when the body is in a physiologically fasted state. Some nutritionists claim a snack under 50 cals won't compromise the outcomes, supposedly because the 50 cals is too small to physiologically break the fast.

This is informally called the "50 calorie rule", and it's relevant because the IF community and nutritionists sometimes endorse it. The point was to communicate that there is a variety of IF protocols beyond what you recognize as fasting.

I will say that it's frustrating for me that you dispute well known concepts and trends in the nutrition community - at least acknowledge what I am saying, and give counter arguments if you disagree. If you can't, I will conclude your motivation here has to do with something other than seeking the truth about our state of understanding of IF. If that's the case, lets end the convo.

Having a meal period <4.5 hours is a form of intermittent fasting called Daily Time-Restricted Feeding. You're fasting 19.5 hours each day.

There are many intermittent fasting protocols including multi-day fasts which have been shown to prolong life in mice. But daily fasting is included in those protocols.

A lot of health hype has focused on Daily Time-Restricted feedings. This kind of IF is typically characterized by a daily meal period of 8 hours or less, with only water or snacks under 50 calories allowed during the rest of the time.

This is supposed to promote changes like reducing insulin resistance, but there are questions about the hype vs truth of this.

None of us are reading it right yet. We're not interpreting the "Hazard Ratios" - whatever those are. Until someone comes along and explains it (or uses chatGPT/bing to explain it), there's no point to form opinions about these results.

In addition, we need to know what percentage of the people were doing fasting, in order to determine if the HRs are statistically significant. That claim a 95% confidence ratio, but is that true?

If we're not able to do this kind of simple checkup on the data, we will just confuse us even more when we find another paper that adds even more naunce and complexity to the issue. That's how people become anti-science. Because we don't understand /this/ work, this work causes us cognitive dissonance. And because we don't understand this work, our dissonance is even worse next time.

I hope a kind statistician comes along and help our thread about by explaining the numbers and what they really mean in terms of death risk and life expectancy.

Can someone interpret the Hazard Ratios for me? What is a HR of 1.3 or 1.8 actually saying?
Re:No Butts About It – A Short Story by Bing Chat on Thursday February 16, 2023 @11:43PM (#34830879)
This was inspired by [1] and [2], two other bing short stories.

First I asked bing to recall Vonnegut's 8 rules [3]. Then I asked bing to write a story about two rival AIs trying to solve a big mystery. In the end, for better or for worse, they will have to work together.

[1] https://news.ycombinator.com/item?id=34829653 [2] https://news.ycombinator.com/item?id=34828153 [3] https://engl270x-2.community.uaf.edu/course-resources/for-fu...