In late June of 2021, GitHub launched a ‘technical preview’ of what they termed GitHub Copilot, described as an ‘AI pair programmer which helps you write better code’. Quite predictably, responses to this announcement varied from glee at the glorious arrival of our code-generating AI overlords, to dismay and predictions of doom and gloom as before long companies would be firing software developers en-masse.
As is usually the case with such controversial topics, neither of these extremes are even remotely close to the truth. In fact, the OpenAI Codex machine learning model which underlies GitHub’s Copilot is derived from OpenAI’s GPT-3 natural language model, and features many of the same stumbles and gaffes which GTP-3 has. So if Codex and with it Copilot isn’t everything it’s cracked up to be, what is the big deal, and why show it at all?
According to a 2020 survey in the US, “organized retail crime” cost retailers $719,548 per $1 billion dollars in revenue. One thief was recently arrested after stealing more than $17,000 worth of power tools from Home Depot. While many stores put high value items in locked display cases, Home Depot felt that this tactic would negatively affect sales, so they partnered with suppliers to add an internal kill switch. Although persistent criminals might find a way to deactivate this feature, it sounds like Home Depot is hoping that will be just enough trouble to convince most criminals to look for easier targets somewhere else.
We would be really interested in getting our hands on one of these power tools to see what this kill switch looks like and how it works. Something like a Bluetooth activated relay is one option, or maybe even something that is integrated directly in the motor controller. If it were up to us, we would probably pick something that receives power wirelessly using a coil and requires a unique code. For their sake, we hope it’s not something that can be deactivated with just a large magnet.
We love seeing old technology brought back to life, especially when it’s done in the context of how the device was originally intended to be used. And double points when it’s space gear, like what [Curious Marc] and his usual merry band of cohorts did when they managed to light up a couple of real Apollo DSKY displays.
The “Display and Keyboard” formed the human interface to the Apollo Guidance Computer, the purpose-built machine that allowed Apollo missions to fly to the Moon, land safely, and return to Earth. Complete DSKYs are hard to come by, but a lucky collector named [Marcel] was able to score a pair of the electroluminescent panels, one a prototype and one a flight-qualified spare. He turned them over to AGC guru [Carl Claunch], who worked out all the details of getting the display working again — a non-trivial task with a device that needs 250 volts at 800 Hertz.
The first third of the video below mostly concerns the backstory of the DSKY displays and the historical aspects of the artifacts; skip to around the 12:30 mark to get into the technical details, including the surprising use of relays to drive the segments of the display. It makes sense once you realize that mid-60s transistors weren’t up to the task, and it must have made the Apollo spacecraft a wonderfully clicky place. We were also intrigued by the clever way the total relay count was kept to a minimum, by realizing that not every combination of segments was valid for each seven-segment display.
Well, this hack has us tickled pink. We love the idea of buying some really cheap piece of technology and doing something amazing with it, and this is a textbook example of that. [davedarko] found the cutest little calculator watch on Ali Express and is working on making a new PCB for it. The plan is to use an ARM processor and Arduino and add a few extras like 24-hour mode and a pink (or potentially RGB) backlight. The new brain will be an ATSAML22G18A, which has an on-board LCD controller and exactly one I/O pin to spare without charlieplexing the buttons.
One of [davedarko]’s primary goals is to keep the LCD and figure out how to talk to it. The first order of business was reverse engineering the watch’s LCD controller by sussing out the secrets from beneath the black blob of epoxy. This was an eye-opening experience as [davedarko] had never worked directly with LCDs before. A strange reading made him bust out the oscilloscope. Long-ish and informative story short, [davedarko] found out that it uses a bias of 1/2 for generating the wave necessary to multiplex the segments and keep the signal alternating. This is definitely one to watch!
We love timepieces around here and have seen all kinds of hacks, especially on Casio watches. Want dark mode? Done. Enable the hidden countdown timer? We’ve got that, too. And have you ever wondered just how water-resistant the F91W is?
All of the technological improvements to vehicles over the past few decades have led to cars and trucks that would seem borderline magical to anyone driving something like a Ford Pinto in the 1970s. Not only are cars much safer due to things like crumple zones, anti-lock brakes, air bags, and compulsory seat belt use, but there’s a wide array of sensors, user interfaces, and computers that also improve the driving experience. At least, until it starts wearing out. The electronic technology in our modern cars can be tricky to replace, but [Aravind] at least was able to replace part of the instrument cluster on his aging (yet still modern) Skoda and improve upon it in the process.
These cars have a recurring problem with the central part of the cluster that includes an LCD display. If replacement parts can even be found, they tend to cost a significant fraction of the value of the car, making them uneconomical for most. [Aravind] found that a 3.5″ color LCD that was already available fit perfectly in the space once the old screen was removed, so from there the next steps were to interface it to the car. These have a CAN bus separated from the main control CAN bus, and the port was easily accessible, so an Arduino with a RTC was obtained to handle the heavy lifting of interfacing with it.
Now, [Aravind] has a new LCD screen in the console that’s fully programmable and potentially longer-lasting than the factory LCD was. There’s also full documentation of the process on the project page as well, for anyone else with a Volkswagen-adjacent car from this era. Either way, it’s a much more economical approach to replacing the module than shelling out the enormous cost of OEM replacement parts. Of course, CAN bus hacks like these are often gateway projects to doing more involved CAN bus projects like turning an entire vehicle into a video game controller.
Amateur radio operators have a saying: When all else fails, there’s ham radio. And that’s true, at least to an extent — knock out the power, tear down the phone lines, and burn up all the satellites in orbit, and there will still be hams talking about politics on 40 meters. The point is, as long as the laws of physics don’t change, hams will figure out a way to send and receive messages. In honor of that fact, the police in the city of Pune in Maharashtra, India, make it a point to exchange messages with their headquarter using Morse code once a week. The idea is to maintain a backup system, in case they can’t get a message through any other way. It’s a good idea, especially since they rotate all their radio operators through the Sunday morning ritual. We can’t imagine that most emergency services dispatchers would be thrilled about learning Morse, though.
Just because you’re a billionaire with a space company doesn’t mean you’re an astronaut. At least that’s the view of the US Federal Aviation Administration, which issued guidelines pretty much while Jeff Bezos and his merry band of cohorts were floating about above the 100-km high Kármán line in a Blue Origin “New Shepard” rocket. The FAA guidelines make it clear that those making the trip need to have actually done something to qualify as an astronaut, by “demonstrated activities during flight that were essential to public safety, or contributed to human space flight safety.” That’s good news to the “Old Shepard”, who clearly was in control of “Freedom 7” during the Mercury program. But the Bezos brothers, teenager Oliver Daemen, and Wally Funk, one of the “Mercury 13” group of women who trained to be NASA astronauts but never got to fly, were really just along for the ride, as the entire flight was automated. It doesn’t take away from the fact that they’ve been to space and you haven’t, of course, but they can’t officially call themselves astronauts. This goes to show that even billionaires can just be ballast too.
Good news, everyone — if you had anything that was being transported aboard the Ever Given, your stuff is almost there. The Suez Canal-occluding container ship finally made it to its original destination in Rotterdam, approximately four months later than originally predicted. After plugging up the vital waterway for six days last March, the ship along with her cargo and her crew were detained in Egypt’s Great Bitter Lake, perhaps the coolest sounding body of water in the world next to the Dead Sea. Legal squabbling ensued at that point, all the while rendering whatever was in the 20,000-odd containers aboard the ship pretty much pointless. We’d imagine that even with continuous power, whatever was in the refrigerated containers must be pretty nasty by now, so there’s probably a lot of logistics and clean-up left to sort out.
I have to admit that I have a weird love of explosive bolts. I don’t know what it is, but the idea of fasteners engineered to fail in a predictable way under the influence of pyrotechnic charges just tickles something in me. I mean, I even wrote a whole article on the subject once. So when I came across this video explaining how the Space Shuttles were held to the launch pad, I really had to watch it. Surprisingly, the most interesting part of this story was not the explosive aspect, but the engineering problem of supporting the massive vehicle on the launch pad. For as graceful as the Shuttles seemed once they got into orbit, they really were ungainly beasts, especially strapped to the external fuel tank and booster. The scale of the eight frangible nuts used to secure the boosters to the pad is just jaw-dropping. We also liked the idea that NASA decided to catch the debris from the explosions in a container filled with sand.
With the principles of molecular biology very much in the zeitgeist these days, we thought it would be handy to provide some sort of visual aid to help our readers understand the complex molecular machines at work deep within each cell of the body. And despite appearances, this film using interpretive dance to explain protein synthesis will teach you everything you need to know.
Now, there are those who go on and on about the weirdness of the 1960s, but as this 1971 film from Stanford shows, the 60s were just a warm-up act for the really weird stuff. The film is a study in contrasts, with the setup being provided by the decidedly un-groovy Paul Berg, a professor of biochemistry who would share the 1980 Nobel Prize in Medicine for his contributions to nucleic acid research. His short sleeves and skinny tie stand in stark contrast to the writhing mass of students capering about on a grassy field, acting out the various macromolecules involved in protein synthesis. Two groups form the subunits of the ribosome, a chain of ballon-headed students act as the messenger RNA (mRNA) that codes for a protein, and little groups standing in for the transfer RNA (tRNA) molecules that carry the amino acids float in and out of the process.
The level of detail, at least as it was understood in 1971, is impressively complete, with soloists representing things like T-factor and the energy-carrying molecule GTP. And while we especially like the puff of smoke representing GTP’s energy transfer, we strongly suspect a lot of other smoke went into this production.
Kitsch aside, and with apologies to Lewis Carroll and his Jabberwock, you’ll be hard-pressed to find a modern animation that captures the process better. True, a more traditional animation might make the mechanistic aspects of translation clearer, but the mimsy gyre and gimble of this dance really emphasize the role random Brownian motion plays in macromolecular processes. And you’ll never see the term “tRNA” and not be able to think of this film.
