I wonder how much of that is because of the way you're using them. They give a lifespan estimate, but that's making some very broad assumptions about how you use them. Those estimates about how many years they'll last are based on you using it for so many hours per day but only turning it on a few times per day. If you turn the light on and off many times per day, as you might in a bathroom or if you're using an occupancy sensor, the filaments will wear out a lot sooner than the projected lifespan. If you're really turning the lights on and off a lot, LEDs are probably a better choice.
Their energy savings is not that much better than CFLs...
That depends on what you consider "much better". The newer LED bulbs at big box retailers like Home Depot are now using around 1/3 less power than equivalent CFLs. That's not the same kind of savings you get from switching from incandescent to CFL, but it's still substantial. If power costs more than about $0.10/kW, they're probably worth the increased up-front cost.
They also seem to have an above-average chance to push management to jump on some new framework bandwagon because they think that will solve all their problems. To be a really good programmer, you have to know how to program, understand the processes that you're automating with your code and realize that no silver bullet will allow you to NOT understand the processes that you're automating. If you don't understand what you're trying to do, you're not going to do it very well.
Sometimes, the cheapest and most efficient LED bulbs are in the blue end of the spectrum, especially when the color temperature doesn't matter too much - like a flashlight.
In that case, it's not so much the color temperature as it is the spectrum. The color temperature tells you what temperature of blackbody radiation your light source most closely resembles, but it doesn't tell you how closely it resembles it. Our eyes work best with light that has a distribution similar to blackbody radiation, i.e. with a wide, smooth distribution of wavelengths. If the distribution has sharp spikes, it can cause things to look the wrong color compared to what they're expected to look like. This is most obvious if you get one of the LED lights that uses a mix of pure red, green, and blue to simulate other colors; you can get something that looks like white if you look directly at the lights, but nothing they shine on looks right. That color shift is what CRI (color rendering index) is supposed to measure.
Lights have to have a CRI of at least 80 to qualify for Energy Star, which means that most household lights are now fairly decent. Cheaper lights and ones not intended for general illumination may go for higher efficiency at the cost of lower CRI, which is what you're probably noticing in the light from flashlights. High CRI (90+) lights are available, but they're usually a bit more expensive and less efficient.
Generally speaking, anything with lots of parts has more points of failures.
Maybe that's true in general, but in the specific case of lighting, incandescent lights obviously have a much shorter life span than CFLs or LEDs. There's plenty of reason to think that incandescent lights do badly with power spikes. My experience is that they're a lot more likely to fail when you turn the light on than any other time, which suggests susceptibility to power surges. It's just that replacing dead incandescent lights is a regular activity, so the occasional failure due to power spikes is much less noticeable than for a light you expect to replace once or twice a decade.
I wouldn't be so sure that energy efficient lights a lot more sensitive to dirty power than incandescent lights. It's just that incandescent lights have such a high background failure rate. If a CFL or LED light dies, you assume there must be a problem with it because their rate of natural death is so low. With incandescent lights, you would have a hard time telling whether one died because of bad power or because it's just given up the ghost.
I find the whole think kind of surprising, since it is known that the whole brain doesn't go to sleep at the same time. Sleepwalking happens when part of it isn't asleep at all.
Neat. Could be used during surgeries instead of anesthesia, or could be weaponized to disable enemy combatants.
Sure, just capture them and subject them to brain surgery for the implant, then turn them lose so you can capture them easier next time.
I have a few. whichever ones i successfully snatched off the playground. usually the slowest runners.
In case you haven't heard, there's a more fun way of getting kids.
Those damb religio-political dogmatists keep blocking publication of my papers on the theory of anturgic phrogneal boropathy.
he believes in some mythical exponential increase in computer intelligence