arafa

From my perspective as a data person, everything on this list is true. I would add on to #4 to say "You're likely the only data person" to say "You're likely the only data person and expected to do everything you need to do your job yourself" (from sourcing the data to deploying your model).

High Data Scientist salaries and expectations combined with a shortage of qualified people often mean you're expected to be a one-person band, which I find to be miserable.

These are ketone ester drinks, but nothing really mainstream or in energy drinks. Esters are far too expensive and the taste far too bad for that. Most people have never even heard of these, in my experience.

I have a fair amount of experience with ketone esters and did some research. They taste terrible, but not as bad as I expected and definitely wasn't a deal breaker for me. I got more used to it over time. They're great for fueling from fat instead of glycogen, especially during exercise, (was very helpful for not having to down a ton of gels during a marathon) and not very helpful for weight loss (as you suggest).

It depends on how fast a caffeine metabolizer you are. I've been classified in genetic testing (whatever the reliability of that is) as a fast metabolizer, and I can have a full cup of coffee 2-3 hours before bed and not stay up. Other folks who I know that got slow metabolizer on the test can't have caffeine after noon without it affecting their sleep (maybe this is the case for you).

The review situation is bad enough on Amazon now that I won't buy any medium to high value product without using ReviewMeta. I find their service to be invaluable. Reading reviews and using ReviewMeta still isn't enough, but it's reasonable and much better than not using them.

"It is all about efficient use of real estate."

I've been a part of these discussions at very high levels and this is the consistent theme. Short-term, easily quantifiable savings against nebulous, hard to quantify harms. Facilities folks always win and get a pat on the back. Any push back about the open office should, in my opinion, start with addressing this basic incentive. Everything about collaboration, open communication, etc. is just window dressing used to justify the moves.

I had a very similar experience with RSI and Dvorak. It was getting pretty bad and many years later, my wrists are holding up great. Maybe I'd have learned Colemak if I learned today, maybe not.

Jump bikes are also safer as the brake lines are a lot harder to cut than the Lime ones (since they're shielded). This happens quite frequently to people I know.

As suggested, it's an excellent but difficult read. But like many such books (say The Beginning of Infinity), they reverberate in my mind for years afterwards, with interesting inferences and callbacks.

That said, you might skip the digression on silver (ugh).

It does seem like a radical option but is surprisingly doable. Beyond Soap was a book recommended here, which was a great read and discourages daily use of soap and shampoo (and conditioner if need be): https://www.amazon.com/Beyond-Soap-Truth-Beautiful-Healthy/d...

I was monitoring indoor CO2 at my home. It definitely swings in the 800-1000 range, especially if I keep the door to the rest of the house closed or mostly closed. But if I leave that door open, it often drops to 600 or less during the day when I'm probably not adding to the CO2. If I open a window it drops to about 430 in a few minutes. It's noisy where I live, though, so I mostly don't ventilate much externally now and live with it.

I went to post this exact comment and you already have. This, at the very least, demands explanation. The book was pretty damning.

My friend used to be into "bending" talking toys this way. My favorite was the bent Furbys they took the "fur" off of: https://www.youtube.com/watch?v=srf1oreJ_Vo

Indeed. The word is that's the new policy and standard (adjustable height desks for all).

I had a professor years ago that emigrated from Bulgaria. He said he could've lived like a king there with relatives in the corrupt train sector/mafia, but came to the US to teach as an adjunct instead. That about sums it up.

Corruption is mentioned in the comments here but I'm disappointed it wasn't mentioned in the article.

Not the first time Goedel's Incompleteness theorem has been related to consciousness. It's a running theme in Douglas Hofstadter's work, especially in Goedel, Escher, Bach.

I think Hofstadter's answer to this question would be: To what extent is AI engaging in and using self-reference? That's the degree of consciousness it has.

https://en.wikipedia.org/wiki/G%C3%B6del,_Escher,_Bach https://en.wikipedia.org/wiki/I_Am_a_Strange_Loop

As a counterpoint (also discussed in other threads below), often few other people want to bid on those keywords, so it's not as expensive. It's still a crappy thing to do.

Reminds me a lot of the ancient idea of Platonic forms: https://en.wikipedia.org/wiki/Theory_of_Forms

Obviously not the same thing, but I think it's an interesting association.

See also this, for more specifics on the math involved: https://effectsizefaq.com/2010/05/31/what-is-statistical-pow...

The basic idea is: Get more data, especially if you're measuring something subtle.

I am pretty surprised to see no discussion of statistical power in the article and very little mention in the comments here. To me, having more statistical power solves many of the issues mentioned in the article. Many of the rest can be handled with use of Bayesian priors, context-specific p-values thresholds (.01, .1, etc.), and replication.

There's decent working guidelines for statistical power, but a lot of the issues (especially around replication) I see are mainly due to sample sizes being too small to adequately detect effects, especially if those effects are small.

They can also be used for money laundering, as a result.

Can't agree more. People always bring up how risky and unethical the insulin pump hacking stuff is in literally every link it's mentioned. I guess intuitively, all things being equal, people feel like like you should be as cautious as possible and test these like any other medical device.

But all things aren't equal. Most of the people doing the hacking (I know a few as well) are already intimately familiar with the biology and consequences, and are already constantly monitoring themselves, or their children and loved ones. They're the best people to test this stuff.

And, like you said, there are compliance issues with children and all sorts of bad things happening already, like devastating lows, diabetic comas, chronic sleep deprivation, etc. The closed-loop and automated systems already reduce these things and improve quality of life for all parties involved. Also the companies are taking forever and may have a vested interest in maintaining their products the way they currently are.

If all things were equal, I could understand the disdain and caution for the device hacking. But they're not.

A U.S. focused take from 2017: https://cliffmass.blogspot.com/2017/02/global-warming-will-b...

Used to happen at our house all the time (once every 6-12 months), we had lots of trees near power lines. There'd be a bang, a blue flash of light, a dead squirrel, and no power (or a power dip).

As someone who uses statistics all the time at work, I sympathize so much with this article and greatly enjoyed it. Every time I try to introduce a Bayesian prior, coworkers either look at me like I'm crazy (because they've never heard of or used Bayesian stats) or like I've suddenly gone soft and introduced a bunch of nebulous, touchy-feely context into the objective truth (if they're dedicated frequentists).

Then we promptly switch back to p-values of .05, a lot of the time not even bothering with a statistical power calculation. I've had better success with introducing power, though. I suspect that's because we can fit it into the existing frequentist framework.

Thanks for the suggestion. Other friends of mine swear by Foobar2K, but I can't stand the interface (a combination of hard for me to use/configure and bad aesthetics), so much so that I still prefer to use Winamp.

Good to know, thanks! I didn't see this.

Seems interesting, especially since I used to use Rockbox. But my iPod Classic/6G is in the Rockbox unstable category and has apparently been that way for 2 years. Bummer, I still use mine all the time. :(

I know folks that work on Nightscout and I agree with everything in this post. I find them to be very thoughtful and circumspect about the work (the contributors often have T1D or family with T1D). Besides that, a closed loop system (which as yet is only partially implemented or is still somewhat inconsistent) is a holy grail for a lot of these folks.

Most of them are already well acquainted with manually managing insulin and the existing products and can handle any mishaps for the most part.

Can't agree more, having been in a very similar position. From what I understand (based on the book SQL for Smarties), it's also more efficient for the query planner in some cases (especially when filter/non-join predicates are added to the JOIN).