The Tesla Model Y car became the best-selling car (not the best-selling electric car, but the best-selling car of any kind) worldwide in 2023, despite its purchase price being significantly higher than the second-best-selling car. And within the electric car subset of the market, Tesla cars have been outselling most other brands of electric car by a very significant margin (I vaguely recall a chart showing Tesla outselling the next 10 or so rival electric car brands combined in the USA). So, yes, I stand by my statement that Tesla cars are considered to be highly desirable by people who can afford them. Your self-righteous attitude that, because you dislike Telsa, people who buy Teslas must be a "rich idiots who doesn't deserve their money" is irrelevant to my statement.

Your criticism that Tesla is often several years late in delivering its pre-announced products is justified, but it is irrelevant to the point I was making, which is that Tesla likes to pre-announce its products by several years while Apple tends to be secretive about its not-yet-released products and then announces them just a few days/weeks before they will be available to purchase.

Here's a quick summary of this thread of discussion. One person claimed that Telsa was very similar to Apple, albeit in a different industry. I politely and partially disagreed, and pointed out several ways in which the two companies are different. My post did not seek to praise Tesla; but rather just to compare and contrast it against Apple. And then you decided to "contribute" by posting irrelevant statements, based on your dislike of Tesla. You are welcome to dislike Tesla, but posting your dislike when it is irrelevant to a discussion reflects badly on you, and makes you look like a troll. Grow up.

I only partially agree. In the sense that both companies have products that are considered to be highly desirable by affluent consumers, yes. But the two companies differ a lot in their business practices. For example, Apple is usually very secretive about not-yet-released products, while Tesla is known to pre-announce new products by several years. Also, although a few decades ago Apple tried licensing its computer operating system to other hardware manufacturers (which gave rise to Apple clone computers), it gave up on that strategy and now only Apple computers can use Apple's operating system. In contrast, Tesla has been making some of its technology available to other EV makers for many years, such as taking an open-source approach to its patent portfolio, recently sending a document containing advice on designing cars with a 48-volt architecture (to replace the less efficient 12-volt industry standard), and for many years inviting other EV manufacturers in America to co-develop supercharger stations (unfortunately, this offer was ignored until recently).

I assume you are referring to the Tesla-proprietary connector being recently standardized as the North American Charging Standard (NACS) connector. Tesla had ignored the American CCS1 standard since it was significantly inferior (it supported much lower charging rates than Tesla's own connector). In Europe, Tesla has been supporting the standardized CCS2 charging connector (which is much better than the American CCS2 standard) for some years.

You can do Google searches to verify that: (1) the number of hydrogen fuel stations is about 1000 times less than the number of EV charging stations; and (2) the number of hydrogen-powered cars is about 1000 times less than the number of EVs. So even if a hydrogen-powered car is a good idea (and I don't think it is), the adoption rate for hydrogen cars and fueling infrastructure is so far behind that of EVs that there is very little chance of hydrogen cars succeeding at becoming mainstream. The following YouTube video provides an interesting overview of some of the significant challenges faced by using hydrogen as a fuel: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

One problem with empirical data is that it can be open to opposing interpretations. For example, I have read the claim numerous times that marginalised groups tend to suffer more mental health problems than the general population. Apparently, there is empirical data to validate that claim. The problem is that I know of two opposing ways to interpret such data: (1) "prejudice against [name of marginalised group] results in them having increased levels of mental health problems, so the high rate of mental health problems is strong evidence that [marginalised group] faces extreme prejudice "; and (2) "higher rates of mental health problems in [marginalised group] is a symptom of such people being genetically inferior".

I can illustrate this with some empirical data provided by you: "in the US, whites have a homicide rate around double that of other developed nations. For blacks, it is around six times higher than whites, of 12X other countries." One way to interpret that data is to claim it shows black people are intrinsically more likely to commit murder than white people. An opposing way to interpret that data is to point out that if white US people are twice as likely to commit murder than white people in other countries, then this shows the likelihood of a white person being a murderer is not due (or, at least, not entirely due) to the colour of their skin, but some other factors are likely to play a significant role, perhaps something to do with the culture/society in which they live. As such, the higher rate of homicide among black US people may well be due to something to do with the culture/society in which they live rather than being something intrinsic to skin colour. Indeed, you pointing out that the homicide rate in the US has decreased significantly over the past 40 years suggests that changes in culture/society during that time period is, at least partially, responsible for the decrease in the homicide rate, which in turn indicates that that the homicide rate is not dictated (solely) by a person's skin colour.

You believe one thing (that black people are intrinsically more likely to be murderers than white people), and I don't share that belief. I am not interested in debating the validity of the belief (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fxkcd.com%2F386%2F). Rather, I was pointing that the argument you used to support your belief was flawed, and that belief bias might have made it difficult for you to see the flaw in the argument you used.

I was not commenting on your critique of maths. Rather, I was challenging your statement: "Can you accept [men are intrinsically more likely to murder than are women] as fact? Then why not for race?" The wording you used asked readers to accept "as fact" that people of one skin colour are intrinsically more likely to murder than people of another skin colour. That is not a "fact". Rather, it is a hypothesis for which you provided no evidence. In essence, I was challenging you on a semantic misuse (fact != hypothesis) that was a flaw in your argument, rather than trying to debate the validity (or otherwise) of your argument's conclusion.

The psychological term for this is "belief bias": https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2F...

Your logic is flawed, because murder rates within a particular demographic vary by location, such as neighbourbood within a city, from one city to another, or from one country to another, and also by poverty/affluence level. As such, the chances of, say, a randomly picked white dude from one neighbourbood/city in America being a murderer is different from the chances of a randomly picked white dude in another neighbourbood/city/country being a murderer. And since (1) there tend to be mostly-white neighbourboods and mostly-black neighbourboods, and (2) income levels vary significantly for different ethic groups, it seems unwise to conclude that people with one skin colour are inherently more likely to be murderers than people with a different skin colour. Certainly, you have not provided any evidence to support such a conclusion.

For people who miss the "There's Something About Mary" reference, here is a link to the original: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

You are being disingenuous. Elon Musk has consistently stated that: (1) he thinks all government incentives should be abolished, not just for electric car companies, but also for ICE car companies and the oil industry; and (2) but as long as their are government incentives being offered, he would be a fool to turn them down. Views (1) and (2) are compatible, and Elon Musk holding both views does not mean he is hypocritical.

Incorrect. The loan in question was not a special loan available only to Tesla. Instead, it was part of a scheme that any company could apply for, and a prerequisite of a company being eligible for receiving such a loan was that the company was profitable. Tesla was profitable at the time. Therefore, Tesla did not want or need the loan to survive, as you assert. Rather, Tesla used the loan to speed up development of its next product (the Model S). Source: Ashlee Vance's warts-and-all biography of Elon Musk.

My recollection of events is very similar to yours, except regarding point 5. I know Musk had been sleeping at the factory during the "Model 3 production hell" ramp up that had almost caused the company to go bankrupt, but that hell had ended the previous year, that is, in 2019, and Tesla had been profitable since Q2 2019. As such, I am not sure that Elon Musk was still sleeping at the factory.

Regarding point 4... At the time, I was surprised that the media's sympathies lay with the bureaucracy that was delaying the re-opening of Tesla's factory, for the simple reason that the initial COVID lockdowns had caused massive damage to the economy, so if a safe way had been found to re-open factories, then surely it was in the economic interests of the country to re-open the factories as soon as possible to help re-start the economy. Yes, I know that Elon Musk is controversial person who is disliked by many people, but an attitude of "If Tesla's factory stays closed longer than necessary and this financially hurts Elon Musk, then that's a bonus" ignored the reality that the longer the factory stayed closed, the longer the employees at the factory stayed out of work and hence suffered financially.

Your portrayal of events is inaccurate. All auto factories had been closed down due to the COVID-19 pandemic. Then all the auto factories except for Tesla's Fremont factory were permitted to re-open, and some bureaucrats seemed to be deliberately delaying giving permission for Tesla's factory to re-open. That's when Musk publicly got annoyed at his Fremont factory being closed. In addition, Tesla already had experience of dealing with a factory closure and then taking safety measures to be permitted to re-open: in their factory in China earlier in the year. Because of this prior experience, Tesla had more experience of what was required to safely re-open a factory than most other companies in the USA, and had a "return to work playbook" that was probably more extensive than "return to work" guidelines that other US companies were using. Here are two relevant documents:

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fcleantechnica.com%2F2020%2F05%2F10%2Fdoes-alameda-county-not-realize-tesla-has-experience-success-reopening-after-coronavirus-pandemic-in-china%2F
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.tesla.com%2Fen_GB%2Fbl...

Like you, I don't agree with Musk's views on COVID vaccination.

You are either ignorant or being disingenuous. Biden went out of his way to ignore Tesla and instead falsely claim that "Detroit", that is, the "Big Three" legacy auto companies, especially General Motors' CEO Mary Barra, had been the ones responsibly for driving the transition from ICE vehicles to electric vehicles. You can see this blatant lie being stated in this video clip: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

That was not an isolated incident displaying how Biden, for whatever reason, was on an anti-Tesla crusade. When faced with such gaslighting from the American president, it is unsurprising that Musk started to sour on him.

Again, you are being inaccurate and disingenuous. The $450 million loan was not to "save" Tesla, since Tesla was a profitable company at the time. Rather, Tesla's first car, the Roader, was on the market and Tesla wanted to raise money to help speed up the development of its next car (the Model S). Unfortunately, fallout from the 2008 recession had made it difficult, if not impossible, to raise money from the open market. Hence, the request for the government loan. The important point was that Tesla did not "need" the loan to survive as you suggest. Instead, it wanted the loan to help expand quickly. Source: Ashlee Vance's warts-and-all biography of Elon Musk.

In essence, testing imperfect autonomous driving vehicles on public roads is the Trolley Problem (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FTrolley_problem), and it boils down to the following trade-off (I'm going to use made-up numbers for illustration):

Testing autonomous vehicles in computer simulations or on privately-owned test tracks maximizes safety during testing, but (it is argued) will result in it taking, say, 20 years to get a fully safe technology. In contrast...

Permitting testing of imperfect autonomous vehicles on public roads will almost certainly results in some deaths due to imperfections in the technology, but (it is argued) this approach enables the gathering of numerous edge-cases that can greatly speed up the development of the technology, so it will take, say, 5 years to get a fully safe technology.

Which is better? Getting a fully-safe technology in 20 years with no imperfect-technology-caused deaths during those 20 years? Or getting a fully-safe technology in 5 years but with, say, 2000 imperfect-technology-caused deaths during those 5 years? This is where the trolley problem kicks in. By developing the fully-safe technology 15 years earlier, you can save hundreds of thousands or millions of lives that would otherwise have been caused by human-driver-error during those 15 years.

If you do an internet search for "BYD innovation" (or just look at the company's Wikipedia page), you will see that your statement is incorrect.

One or two years ago, the most powerful RISC-V core offered a fraction of the performance of the cores in a Raspberry Pi 4. Today, the most powerful RISC-V core offers performance similar to the cores in a Raspberry Pi 4 (but there is a CPU containing 64 of those cores for companies wanting to experiment with server-grade computers based on RISC-V). Within a few years, I expect RISC-V cores to offer laptop/desktop-level performance, and a few years after that RISC-V chips will offer data center-level performance. Put simply, RISC-V may not be competing against Intel/AMD or ARM today, but it probably will be before the end of this decade.

A few months ago I read an excellent, slim book called "The RISC-V Reader" by David Patterson and Andrew Waterman. The book addressed two topics and intermingled discussion of both topics within each chapter. One topic was a tutorial on RISC-V assembly language, which was of little interest to me since I am not an assembly language programmer. The other, and unexpectedly fascinating, topic discussed the "why" behind architectural choices in RISC-V. One takeaway I got from this "why" topic is that history has shown each CPU family (x86, ARM, MIPS, 68000, Sparc, PowerPC and so on) undergoes changes from one generation to the next. Typically a new generation offers more instructions than the previous generation, but occasionally a new generation will remove support for a capability provided in an older generation, because the chip designers learn lessons of the form, "[such-and-such] seemed like a good idea at the time, but with the benefit of hindsight we recognize it was a mistake so we are removing that capability from the new generation of our CPU". The designers of RISC-V studied many such lessons from existing CPU families and used those lessons to make better design decisions for RISC-V. In essence, the RISC-V designers have been able to do an excellent job because they have been standing on the shoulders of giants.

The base instruction set for RISC-V consists of just 47 instructions. A CPU implementing that base set might consist of as little as a few tens of thousands of transistors, which means it will have a high fabrication yield and be very cheap to manufacture. Such RISC-V chips will be able to compete effectively in the embedded systems market. That base set of instructions lacks support for floating point operations, integer multiply and divide, some bit manipulation instructions and instructions for niche areas, such as vector arithmetic or cryptography. Optional extension packages provide those instructions, and you can expect a future smartphone/laptop/desktop/server-grade RISC-V CPU to provide them. Even with such optional extension packages implemented, the number of instructions in RISC-V is a tiny fraction of the number of instructions in, say, x64 or ARM64. Thus, a powerful RISC-V core will require far fewer transistors to implement its instruction set than are required to implement the far larger instruction sets in x64 or ARM64, and thus the RISC-V core will have a larger budget of transistors for implementing performance-enhancing features such as large caches, branch prediction, and more instruction-level parallelism. At least, that's the message I got from the book. If that message becomes reality, then RISC-V will be able to compete effectively against the incumbent market leaders.

I have presented a counter-argument in another comment on this post: https://f6ffb3fa-34ce-43c1-939d-77e64deb3c0c.atarimworker.io/comments....

I agree with the response by Firethorn, but I want to put some made-up numbers on it to illustrate the argument...

Let's assume there are 40,000 road fatalities in the US annually, and let's assume 90% of those are caused by driver error (the other 10% are presumably caused by, say, mechanical failure or acts of god). So that is 36,000 preventable deaths per year that might be avoided if/when autonomous driving is perfected. Now let's assume there are two potential options for achieving perfect autonomous driving.

Option 1 is to test a (not-yet-perfect) autonomous driving system in a lab for, say, 20 years until it is perfected. Then it can be used publicly. This approach will result in zero road fatalities from the not-yet-perfect autonomous driving system.

Option 2 is to permit the testing of not-yet-perfect autonomous driving systems on public roads. Let's assume this approach will shorten the time until a perfect autonomous driving system is achieved from 20 years down to, 15 years, and also assume that this public testing results in an average of 100 extra deaths per year until perfection is achieved.

The cost of option 2 is an extra 100 deaths annually for 15 years = 1500 extra deaths in total. That is bad. However, by knocking off 5 years from the time required to get a perfect autonomous driving system, it saves 36,000 deaths annually during each of those 5 years, thus saving 180,000 lives. The possibility of sacrificing 1500 lives to save 180,000 lives is an example of the "trolley problem" (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FTrolley_problem). Some people may not be happy with such a "sacrifice some lives to save many more lives" decision, but some other people are happy with it, and this is why it can be argued that permitting the testing of imperfect autonomous driving on public roads is beneficial overall.