Posts by Anonymous Cowpilot 53 publicly visible posts • joined 31 May 2023
I am not sure you really understand the complexity, cost and risk in long haul flying. It's optimised for safety not comfort.
The Singapore Airlines magic dust is thst they fire any cabin crew that so much as frown in front of a passenger for a few seconds. They have a reputation as a brutal employer for both cabin crew and flight crew.
They aren't talking about replacing the air/ground radios with IP based equipment, they are talking about replacing the network that connects rafars, tracons and towers together with network infrastructure that people know how to maintain and you can get spare parts for.
We did that in the UK decades ago (2002).
Indeed, but unfortunately the UK CAA (as part of the post brexit rush to distance from anything European) has a policy of distancing itself from EASA and aligning more closely with the FAA. Or at least had that policy under the last government and has not got round to changing it yet.
The policy extends to preventing aircraft engineers with EASA certification from getting equivalent CAA certificates and working on UK registered aircraft. Instead they must retrain from scratch, despite a lack of qualified engineers crippling the aviation industry at the moment.
Indeed, while not a specific exclusion for helicopters, the "land clear" flight rule requires that (other for exempt flights) you are always in a position to land in an unpipulated area if one of your engines fails. For a single engine aircraft this means in a glide or autorotate. This is relatively easy for planes that mostly have glide ratios of 9:1 or above, but not for helicopters. In practice it's only an issue over very large cities even for helicopters, but a self flying hover taxi isn't going to get much use over a small city.
London has class D over the whole area, but even if you get permission to enter the class D (which is very rare but there were a few people that did it during COVID when there was very little traffic) you still need to land clear and they don't count the Thames as clear because it's a busy waterway. The pilots who did fly single engine over London during COVID got calls from the CAA reminding them of the land clear rule (and that the Thames doesn't count, which is not obvious).
This is indeed odd, although amy just be to avoid having to pay 30% of any revenue they make through the apps to Google/Apple.
The concern with the drones though, is not that they currently contain spyware or backdoors, but that should DJI be required to by the Chinese government, it would push such features as an OTA update, giving the chinese government an army of drones that if nothing else, could be forced to fly into restricted airspace, severely limiting the capabilities of the US Airforce. While such scenarios seem like hollywood conspiracies, I would have said the same about exploding hundreds of pagers and walkie talkies 6 months ago.
One of the features of almost all cryptocurrencies is transparency - all currency movements are on the ledger, so knowing which wallets have lots of currency in them at any time is simple. Its a little harder to match wallet addresses to real people, but some people (especially those that seem to think that cryptocurrencies are anonymous) use "vanity addresses" which make it much simpler.
They are not just emergency decompression suits - they are part of the spacecraft. They contain all the sensors that detect the wearers heart rate and other vital signs, together with compute systems and software that request coolant etc from the space capsule to keep the wearer "comfortable" and provide the necessary G-suit style constrictions to manage blood flow around the body (if required). They are much closer to the suits worn by fighter pilots (which are also unique to each aircraft type) than to pressure suits.
And more importantly, an investigation can determine if there are things that can be done to prevent such tragedies in the future. For example are there design or manufacturing flaws that may be present on other vessels. There may also be cultural changes or reinforcement (for example did the employer/employee relationship and desires to deliver top customer service make the crew defer to the customer's about the need to have hatches open due to the hot weather.
I am not familiar with sailing, but in aviation pilots being employees of their passengers has been the cause of multiple accidents. Aviation law has been updated to be very clear that the captain is the ultimate authority on an aircraft, independent of factors such as employee/employer relationships, seniority, etc.
Its absolutley a question of safety - all of aviation is about safety and for good reason. There is no good reason why drone pilots should not have to follow the same restrictions I do as a pilot, and no reason drone manufacturers should get away with lower saftety levels compared to plane or helicopter manufacturers.
In my view the regulations are already too loose, there have been multiple near misses where people have been flying recreational drones in the circuit (less so recently since rules were tightened) at the airport I fly from.The only option to ATC at that point is to close the airport, which is a huge pain.
My understanding is that they are closer to the 6ft size so are likely to have MODE-S and ADS-B, and the trial uses"special use airspace" (SUAS) for flying. My understanding is that the whole purpose of renaming "Danger Areas" as "Special Use Airspace" was for this purpose, and my guess is that the SUAS area around Fairford (DA218A/B) is part fo this trial.
This is not a mixed use airpsace trial. VFR traffic will be prohibited from entering the SUAS while the drones are flying.
The suits aren't just clothing, they have all sorts of sensors that measure astronaut vital signs and feed them to the ship for communication back to earth, as well as local processors that use the vital signs data to adjust environmental parameters (and request appropriate coolant etc from the ship). The suits have a lot of networking and processing going on. It's more like trying to connect a tesla to a BMW diagnostic station than swapping between jumpsuits with different logos.
Studios use human actors because they bring a level of realism and emotion to the role that AI cannot yet replicate, and is probably still quite far from achieving. If studios (especially smaller ones) want to forgo actors for AI that is fine - its likely they would not have hired actors originally anyway. However, once an actor has done most of the hard work to create a character and make them feel human, its relatively easy for an AI to take that and interpolate additional scenes, which is what the actors are hoping to avoid.
One reasonable analogy is audio books - good text-to-speech has been around for decades and is a built in feature of most e-reader software. However there is still a large (and growning) market for audio books - because actors bring emotion in a way that even the best text-to-speech engines cannot. While AI can do some interesting stuff, I fully expect "created by humans" to be seen as a mark of high-quality products once all the hype has died down. We will be willing to pay more for products where we can feel human emotion and creativity from them.
While the claims are not as outlandish of those of McKinsey and company, the paper is pretty bad. The authors seem to share the view that is common among academics, that software engineers wotk on code from the start to the end of their day. Recent studies show software engineers spend between 20% and 50% of their time on code-related tasks. Any tool that claims more than a 20% productivity boost needs to be looking outside "programming" to the wider engineering aspects.
If managers want a 50% improvement in software engineer productivity a good start would be getting rid of management meetings, HR training courses, excessive time management processes and random "leadership initiatives"
It has always been required that you need a re-entry license for re-entry, but as mentioned in the article, you did not have to have that license at the point of launch. This left the situation where you could legally launch an object that is due to come back to earth without yet having the license to bring it back. In the Varda case they just left the re-entry vehicle in orbit until they had the necessary license and then brought it down. Quite what would happen if you were denied the re-entry license is not clear, as eventually the orbit would decay and the re-entry vehicle would come down no matter what the FAA says.
What the FAA has doen is close the loophole so that you can't launch now and then hope to get your re-entry license before you run out of fuel. You now need the re-entry license before launch.
Something similar to what is described could be a misremembered scenario. As you say, deltree deletes everything under the specific directory so deltree c:\windows would do a reasonable job of removing windows. However as a dos utility it wouldn't follow windows shortcuts and would instead delete them, and windows didnt support real links so I can't see a scenario where it would escape from the windows directory and delete the whole disk. Chckdisk would help clean up lost sectors and mismatches between the FAT and what is on the disk, but would not do anything clever with directory trees.
Most likely to cause the described behaviours would be if the person ran deltree c:\ instead of c:\windows
That sounds eminently sensible so I think what is much more likely to happen is that His Majesties Government forces BT to spin out the part that owns the poles into a separate publicly traded company with a monopoly on poles. This company then has to maximize shareholder value by making 300% markup every time they install a pole, and must send 3 managers and a safety supervisor to observe every time someone wants to climb a pole. It would be a coincidence if the largest shareholders in British Poles Ltd turn out to be donors to the sitting government.
That sounds much better than around here, where CityFibre have recently been along my street adding a new pole 5ft away from each BT pole. Most of the BT poles only appear to have a handful of cables on them.
So in our street we can choose between BT fibre and CityFibre each using a separate set of poles or virgin fibre from underground.
They don't know if it posed a danger to anyone on the ground, but this is why they have to report it. Any item falling from an aircraft or part of an aircraft becoming detached during flight has to be reported to the local aviation authority as an incident specifically because it could injure someone or damage something.
Helicopter pilots are told to do that. Plane pilots are told not to do that. We plot a course on a map and adjust for wind speed and compass error. Then we fly the course and check way points to see where we are.
This is still the main approach to navigation for light aircraft and the only method allowed in an exam.
We already get alerts when GPS conflicts with inertial nav systems. It happens a few times on a normal flight so is pretty common. Navigation without GPS is common and pretty easy. However, drones often rely exclusively on GPS and some of this is aimed at the growing use of drones in aviation. Also, recently more and more airports are using approach procedures such as RNAV that need a strong GPS fix. If you don't have a good fix you still know where you are but the airport may not let you join the approach (because it means they have to switch from RNAV to ILS or another approach type that requires much bigger separation between aircraft, which makes the airport less money).
I studied a computer science degree 25 years ago and while it had no specific security module, security was prevalent throughout the course. We learned about separating data and opcode memory and the perils of pre-emptive execution in CPUs in hardware modules. We learned about buffer over and under flow attacks in programming courses and learned about SQL injection in database modules.
And I think this is how it should be. Security should not be a separate module that you can take or avoid, it's not a separate dicipline that you apply after the fact. Security pervades computing and should be incorporated into all aspects of its teaching, not separated out.
Helicopters still have to taxi to a runway or helipad; they aren't generally allowed to just go up from where they are because they may be under an approach track. Most helicopters will hover-taxi, but some will ground-taxi. However having a helicopter take off vertically is incredibly inefficient. Having an e-VTOL take off vertically is the most fuel-efficient route, so this is one of the things they need to figure out for "vertiports".
An e-VTOL that loses a rotor is harder to control than a helicopter that loses its main rotor because a helicopter in autorotation is still subject to fairly balanced forces, whereas the rotors of the e-VTOL are no longer balanced (it will pull in the direction opposite to the missing rotor and will be very difficult to get it to go in any other direction, including straight down.
It's mostly a plea for budget from the CAA - but helicopters and VTOL drones have very different performance and flight profiles, including single rotor failure, as you indicate, but also including how they turn and the most energy-efficient paths for taxi, take-off, and landing. There is certainly some work to do here, if for no other reason than to make politicians and those responsible for legislating and budgeting for these things aware of the (much more significant than people realize) differences.
Given reports say that P2's headset was ripped out of the socket and sucked into the cabin it probably goes something like:
P1: "white with no sugar please"
P2: "hey, that pressurization warning is on again"
P1: "ok, I will get the checklist out, you have con..."
<whoosh>
<crackle buzz> (as the headset gets ripped out of the scoket)
P1: I have control
P1: I can't hear you and I assume you can't hear me, I'm going to turn back and declare an emergency
<rest of communication done by pointing and gesturing>
More seriously they probably want to know if the pilots noticed a pressurization warning some period of time before the depressurization incident. They will know if the warning was made but knowing if the pilots noticed and understood it is important in preventing such incidents in future. The time between a warning being issued by the aircraft, the pilots noticing it and then the pilots working out what to do about it can often be a factor, and redesigning warnings, giving them earlier or having clearer checklists for what to do with a specific warning can often be outcomes of these types of investigations.
My guess is that Coinbase, being all Web3 etc etc, probably has a large number of "contributors" (of content, code, legal advice, etc etc) that they treat as "self employed", have never met, and pay in crypto tokens - leading to a situation where they don't actually know who all the people "working for them" are. This isn't a very sensible situation, but its the kind of thing that the web3 crowd think is cool.
LLMs are not capable of independently providing sourcing - its just not how the model works. However, it is possible to parse a generated response against known training data to provide some sourcing for parts of it. The post-processing approach is much more effective than asking the LLM to generate sources in its text - which tends to result in it "generating" links to unrelated content.
It sounds like it was mostly "small company board politics" from a company that forgot its centre of the world stage. It sounds like the board felt snubbed because Sam Altman was doing something they didn't approve of (theories vary from he was in talks with others about an AI chip startup to that he was moving too fast in trying to improve the GPT models to that the board wanted more focus on AGI and less on LLMs). Rather than deal with this sensibly, the board seemed to forget they are a very visible company with investors and tried to throw their weight around. In most companies with 750 employees, no-one would even notice if the board ousted the CEO, but OpenAI is not most companies and the board seemed not to understand that.
They aren't quite helicopters. Their controls are very different. and they are completely dependent on a computer interpreting a humans inputs to make the necessary changes to engines and control surfaces. That is quite different from most helicopters which are still usually driven by mechanical linkages between the controls and egine/rotors/surfaces.
Taxed in what jurisdiction though? For multinationals its not easy to say "where" a profit was made. If a web store based in luxembourg sells a phone made in china to a us specification running software written in poland by an engineer working for a german subcontractor - who generated the profit?
I agree the laws need fixing, but unless countries with preferential tax laws like ireland come into line its not that simple to solve. Companies just declare that the "profit making" parts of rhe supply chain are in the low tax counties and everything else is a loss making subsidiary.
>>> Anyway, its kind of wrong to use a person's likeness without their permission unless it falls under 'fair use' or that person's long deceased (and even then there's a small matter of the ethics of bringing dead people back to life). <<<
Bearing in mind that fair use is not a universal right. For example, the UK does not have a fair use clause for copyright - it has a much more restrictive fair dealings clause.
But devs are not interchangeable; the skills and knowledge needed for writing game engines is very different from the skills and knowledge needed to write games. Back in my consulting days, I would regularly work with clients that used open source because they felt they could just patch any bugs themselves if the community could not or would not, only to find that their developers couldn't understand the code, let alone patch it.
"radioactivity" (gamma radiation) is also electromagnetic radiation (at the other end of the electromagnetic spectrum). I am fairly sure you weren't thinking that the phones emit large numbers of beta particles (they will emit small amounts of beta particles because the Carbon-14 isotope is present in many organic materials).
Indeed, and you can file a flight plan to many airports that aren't customs airports so don't have IATA codes but do have ICAO ones. You can also fly IFR to private strips that don't have either an IATA or ICAO code. I find it unlikely that an airport name change has anything to do with it.
It isn't retrained that fast because training is incredibly expensive, it was trained on Github in 2021. Any changes in response are either a) just part of the non-deterministic nature of the models (its not clear from the report that they asked multiple times and checked that the model gave a consistent response at a point in time) or b) the result of fine-tuning and updating the prompting of the model, which can be used to introduce specific new facts or point answers in certain directions but isn't a way to incorporate large amounts of random Github code.
An LLM is not a logic machine, its just generating text. When they ask ChatGPT whether 17077 is prime and give it the steps to do so it just generates text of the type that it normally sees when people write down what they have done. Almost every number is not prime, so almost every time they do the exercise they write that it is not a prime. This paper makes the mistake of thinking that if you ask for ChatGPT to do something in steps that it executes a set of steps each time going back to some logic machine - its just generating text. Its the equivalent of asking a student to calculate whether 17077 is prime and show their working and they just find an example online and copy and paste it.
ChatGPT is a chat interface to a large language model - its just generating text, its not doing maths. Unless you ask the steps as multiple prompts then it doesn't decompose a problem to smaller steps and work on each, it just generates the type of text people write when they have done that.
Aviation uses the much simpler approach of VHF (121.5) and UHF (243.0) radios for emergency (and most other) communication. We don't rely on satellite phones (although we will use them for supplemental communication like arguing with company about the overtime implications of a diversion).
When my previous employer (a small business) decided to close the (profitable) business down and use the money in its bank account to buy a villa in Marbella (after the brexit vote but before the laws changed) all 16 employees found out by receiving an outlook calendar invites for "Redundancy consultation #1", "Redundancy consultation #2", "Redundancy consultation #3". None of the invites had any text in the body, just the title.
