

But does the GPS work? Is it able to correctly locate your car on the map? There are other cool things we use GPS satellites for, but their purpose is to calculate the position on a globe. A Global Positioning System, if you will.


But does the GPS work? Is it able to correctly locate your car on the map? There are other cool things we use GPS satellites for, but their purpose is to calculate the position on a globe. A Global Positioning System, if you will.


I think it shouldn’t matter. IIRC, the GPS algorithm depends on time differences, not absolute time. As long as the satellites are coordinated with each other, receivers won’t care which parameters the absolute time matches.


I’ve had good luck with locally connected devices (zwave and zigbee usually). I have a bunch of light switches, temperature sensors, etc. But I bet local connected stuff is hard to find for appliances.


IMHO, the real problem is that it sounds just as “confident” whether the answer is accurate or not.


Sure, but that’s not the situation in the article. These are publicly owned/contracted surveillance cameras that are used to track the movement of residents. And the fig leaf the department is using is to say “the cameras don’t track people” because that’s just something that Flock does later…with data from the cameras.


This has such “guns don’t kill people” vibes.


What about using a variable directly (without the @<(echo ...)? Does that still leak?


The easiest damages would be the cost of attending (conference tickets, travel, hotel). If there was reputational damage, that’s harder to prove but the conference could be liable for that. Seeing how many vocal supporters they have, reputational damage would be hard to prove


Not bullshit. Just another poorly written popsci article promoting real research.


From skimming, it sounds like they’re trying to use compressive sensing techniques, but push the “compute” to a physical, optical structure. That gives you a smaller device without the expensive compute (or the concern about losing data from random noise).
In general, it’s not hard to make a basic optical spectrometer. Most people have seen a prism splitting light into a rainbow. Imagine that plus an array of light sensing pixels. The light intensity on the array is your spectrum reading. The further away you put your pixel array, the more spread apart the colors in your rainbow, but the less light hits each individual pixel.
Optical spectrometers generally use diffraction gratings instead of prisms, but the trade-offs are the same. Longer optical path -> more spectral resolution -> more expensive light sensors.
Compressive sensing tries to break that trade-off by using math from information theory to get a usable data from fewer measurements. The single pixel camera is a great intro to the field. You use a single photodiode plus a series of known masks to take a series of measurements. From the masks + single pixel measurements, you can reconstruct the original image. There’s probably code out there to do it virtually if anyone’s interested. IIRC, to do a virtual measurement, you apply the mask to your image, then sum up the values. The reconstruction process is then identical to if you had real measurements.
For the $10 spectrometer, it sounds like they’re pushing some of that “compute” to a tunable optical system. In other words, the device “takes compressive sensing measurements,” but it does some of the reconstruction before it hits the light sensor.


The writing in the linked article is a little weird, as others have pointed out. However the journal publication is very cool. If this is reproducible, it’ll likely have a noticeable impact over the next 10 years or so. We won’t see $10 spectrometers, but we might see handheld ones in a similar format to the cheap IR cameras.
Here’s the link for anyone interested.


I’ve worked with spectrometers and on spectrometry design before, and that quote from the article is intelligible. It’s also pretentious and poorly explained.


One small correction. You switched units. You started with Watt-hours (kWh, energy) and then switched to Watts (GW, power). With the right units, it’s even more dramatic.
There are an average of about 730 hours in a month. If a home consumes 1000 kWh per month, that’s an average of 1.3 kW. If we divide 9 GW by 1.3kW, we get 6.9 million.
So this data center will use the same amount of energy as over 6 million homes. For reference, Utah has a population of 3.5 million (total people, not total number of homes).
Here’s another way of comparing the numbers. If this new data center uses 9 GW of power 24/7, that’s an about 6,500 GWh per month, or a little under 79,000 GWh per year.
In 2025, Utah produced a new record of over 35,000 GWh.
So this data center would more than triple the amount of energy produced in 2025.


I largely agree, but one situation I can think of where condensing the work makes sense is experienced professionals who already meet the learning outcomes. Their goal is to prove that they know the material, then have a degree to show as proof, not to actually learn the material.


Maradona with an eye patch and peg leg.
For a company “on-prem” might not mean the same as “local”. You might run them in your company data center or in a colocated server (rented space in a data center). In either of those cases, the data center might be physically close to you, but it could be on a different continent.
“On-prem” implies that the company has physical access/control of the server.