I knew the Titanic was discovered in the mid 80s, but was pretty sure it was not not by this submersible, but a ship pulling a camera. So that took the entire sentence into question and that could IMHO very well been an AI that merged nuclear sub in the 60s and Titanic in the 80s because they were both deep sea operations by this vessel or something.
this is the kind of error humans make because we generate text letter by letter. LLMs generate text in whole words or meaningful fragments of whole words (tokens). unless 1986 is unusually statistically similar to 1968, an LLM is unlikely to make that mistake.
"less than 2 drinks a day" would be called "moderate drinking" in my culture. I mean if someone is having a drink or two everyday that is moderate drinking already.
Light drinking for us would be like 3 drinks in a week! 4 drinks/week or more gets us into the moderate drinking territory.
To be fair, "less than 2 drinks a day" covers any range below the upper limit of 2 a day, and hence also includes what you consider light drinking in your culture.
I suppose that the root of the matter is whether or not the range '>0 to <2 drinks a day' on average is too broad and may be hiding a significant difference of outcomes between the top of the range at 2 a day and the (non-abstinents) at the bottom who drink an average of, say, 2 drinks a week.
And that's not even going into regularity. I'm sure that 2 drinks a day on the regular has different health impacts and social implications than 14 drinks a night only on Fridays.
In France, where I am, awareness campaigns recommend sticking to a maximum of 2 glasses per day, for men, 1.5 for women. and not every day.
And these values only take into account protection against serious health problems, such as stroke, not the problems of moderate daily alcohol consumption, such as for example the fact that alcohol being a depressant, consumption will have a impact on their general well-being.
To be honest, I'm even more confused on what to think about this paper.
The J-curve charts I've seen indicated lowest all-cause mortality at 3 drinks / week for men and 2 drinks / week for women. So even if you believe there's something healthy to alcohol (and it's not just the unhealthy people -> have to drink 0), it probably won't be any healthier going past 3 or 2 / week. Portion size is 12oz 5% beer, 5 oz 12% wine, 1.5oz 40% liquor.
Agreed. From what I know, the positive points of certain alcohols such as wine which would protect against heart problems have either been denied, or the same advantages have been found in everyday non-alcoholic products (green tea if I remember well).
This is a general misconception. Alcohol being a depressant doesn't mean that it causes depression - it means that it depresses (that is - it lowers) the activity of nervous system. The opposite of a depressant is a stimulant (like amphetamine), not an antidepressant.
> One such piece of magic new to 802.11n Wi-Fi is called "beamfoming"
That's not quite true. 802.1ln has MIMO (Multiple-Input Multiple-Output) processing, with "multiple" referring to the number of receiver and transmitter antennas. Beamforming is a special case of MIMO, and MIMO is a generalisation of beamforming.
In a "Line-of-Sight" channel with no reflectors, MIMO converges to a beamforming solution. Capacity is then limited by the ability for the rx/tx array to resolve each antenna in the tx/rz array: the diffraction limit.
In a "rich" channel, with reflectors, MIMO converges to a more complex solution, which takes advantage of the angular separation of the reflectors to resolve the individual rx/tx antennas, even if they are too close to each other to resolve with beamforming. Yes, counterintuitively MIMO capacity goes up as the channel become more complex/rich and less line-of-sight, whereas with just beamforming the capacity would typically go down.
You can sort of think of MIMO as being beamforming where beams are bouncing off widely spaced reflectors, but even that doesn't do it justice. In reality, each "beam" is replaced with complex wavefront ("mode") which is matched to the environment and each mode is orthogonal to the other.
They are describing a situation in which a line-of-sight channel is replaced with a rich/complex channel: the exact conditions under which MIMO distinguishes itself from beamforming.
I'd say incomplete rather than incorrect, and the complete story is worth knowing as it makes the solution used more interesting.
(Author here.) Yeah, I get where you're coming from. Ultimately, this was an editing decision first and foremost.
Beamforming is cool and magical, and MIMO even more so. The post wasn't intended as a primer on wireless technology, just as a fun read for folks to enjoy. I tried to sprinkle in some nerd-snipe-quality technical detail and offer links for folks who might want to dig in, and MIMO is explicitly discussed in both the 802.11n and in the several links on beamforming I provided.
I barely managed to explain beamforming without that sidenote turning into a paragraph of its own. I don't think I could have done MIMO justice in a sidenote.
I felt the need to mention the difference as so many people equate beamforming with MIMO and claim that there is no difference, when in fact that difference, which Foschini discovered in 1996, is responsible for spatial multiplexing and the high WiFi data rates we enjoy. I figured the typical HN reader might be more technical and interested in the difference, or maybe not.
Haha sadly from reading many of the rest of the comments, the typical HN reader seems unconvinced that trees or rain could impede Wi-Fi signals, because their Wi-Fi at home goes through walls just fine. A few even suggested I shouldn't have bothered with Wi-Fi and instead just laid cable across a few city blocks instead :)
But I do appreciate precise language and the desire to help people learn new things, so thanks for helping make the distinction between MIMO and beamforming clear! I hope at least a few more people know about it now thanks to your comments.
"A magpie had learned to use the name of a dog on a property in New South Wales. The owners also had a cat that had tried everything to get rid of the magpie. When the cat approached, the magpie did not fly away but called out the name of the dog; the dog came running and chased the cat away. Calling the dog was not mimicry any longer but a most effective way to use the dog's name."
Latency is probably the crucial specification here. LTE includes low latency IoT modes, but last I heard getting them to work was an active area of research. Maybe DECT-2020 is the plan B?
> divide that bandwidth by frequency (like AM and FM) or time
Ah. The real magic is when we separate by space (beyond just frequency or time). The ability to do this was discovered relatively recently, in 1996, by a guy called Foschini, though radio astronomers will say "Meh". By adding multiple antennas and doing space-time coding engineers found they could pump an order of magnitude more data through a radio channel. The maths involved is high school level (linear simultaneous equations), and it's magic to understand Foschini's work and think "Why didn't we do that before?"
The other bit of radio magic is error control coding. This is the stuff that lets us reliably talk to Voyagers I and II.
Fascinating how we keep being inspired by fundamental physics and astronomy to keep cramming mode information in our channels. I'm still trying to understand Orbital Angular Momentum multiplexing https://en.m.wikipedia.org/wiki/Orbital_angular_momentum_mul...
I'd agree with the Wikipedia article, that it sounds like MIMO, in that it requires the beam to have a spatial extent.
From the Wikipedia article:
> can thus access a potentially unbounded set of states
That's what people originally thought about MIMO. MIMO's not unbounded. The limit to the number of states is related to the surface area of the volume enclosing the antenna, with the unit of distance being the wavelength. A result radio astronomers already knew when the comms people derived it. With absolutely no evidence to back it up, I'd guess that the same limit applies to OAM multiplexing.
As an aside, when one expresses physics in terms of information theory my understanding is that the maximum the number of bits that can be stored in a volume of space (also the number of bits requited to completely describe that volume of space) is related to the surface area of the volume with the linear unit being Plank lengths. Is MIMO capacity in some way a fundamental limit in communications?
It's potentially using a bunch of Matlab toolboxes, being a communications related tool, but it's always worth trying to run Matlab code in GNU Octave.
That should be 1986?