
It’s a little less disorienting than when all one had was telephone calls. And I guess that was less disorienting than when all one had was letters and rumours.

It’s a little less disorienting than when all one had was telephone calls. And I guess that was less disorienting than when all one had was letters and rumours.


According to Are Song Lyrics Copyrighted? How the Law Works, unless their use is ‘fair use’ or they have a license, then they are violating copyright, if I understand the article correctly. I believe that site explains laws in the United States. It probably varies somewhat by jurisdiction, so I expect it would depend on who owns the website and where they are based.


Is ChatGPT a legal entity competent to violate copyright law? I don’t think that’s likely.
I do think OpenAI violated copyright law by copying song lyrics and other media to use them as input to their LLM systems, to embed the essence of them into their LLMs for commercial benefit. Judging by the valuations of the companies that do not yet have significant income compared to the investments, on the face of it, the IP they copied, often without license, as far as I know, is fantastically valuable.


Actually, that makes a lot of sense…
Given that Trump is the best president ever (I’m sure I have heard people say so), then anyone who opposes him is, ipso facto, an enemy of the state and should be treated as such.
The more I think about it, the more obvious it is. I mean, why even have another election?

Broken link for me:
The requested URL was rejected. Please consult with your administrator.
Your support ID is ******************************************
Error 403 - Forbidden
F5 site: sv10-sjc
Maybe forbidden for IPs outside CA?
But this reminds me of the recent ruling in the UK:
Technology: UK judge warns lawyers about risks of AI use in court
In my opinion, given all the information available about how unreliable LLMs are, anyone using AI professionally without thorough validation of the outputs used should be found negligent.
For example OpenAI’s recent Why language models hallucinate - which reports an accuracy rate below 25% for two of their models. But there have been many other reports from diverse sources, for a long time. And, in my experience, it only takes a few minutes of interaction to experience significantly false output.


Pulsing the power every 3 to 5 seconds would be fantastic! One pulse every 30 seconds is not good. But I haven’t yet found a manual or sales person that can tell me the period of the pulse width modulation for any brand/model. Not that I have tried very hard.
Maybe I should try KitchenAid. What you describe sounds wonderful.


That was exactly the problem with simmering anything.
Also, only 9 power levels wasn’t enough. It was very powerful (nice when I wanted full power) but the steps in power were too big. For many things the only options were too hot or too cold.
But the fault that made me replace it was an intermittent one: occasionally (about five times in 18 months) it went to full power. This could happen at any power setting. No change in the indicated setting, but the power would come on continuously. Anything other than a pot of water would, in just a few seconds, be burning. Very dangerous! Fortunately, it never happened when I wasn’t standing right there to turn the power off at the wall switch. Being intermittent, technicians couldn’t find/fix the fault. It also occasionally stopped heating for a minute or so at a time, as if there was some thermal lockout even when nothing was unusually hot, but at least that wasn’t dangerous.


It was Haier. And I see they are no longer the biggest by revenue, but still #3
The problem with simmering is that the cooktop was very powerful (nice when you want to heat something quickly) but it only had two modes: one or off. The power was regulated by turning it on briefly, once every 30 seconds. Even at the lowest possible setting (there were 9 power levels), a pot of water would boil each time the power came on for about 3 seconds. Then it would cool for 27 seconds. Even a pot with a thick base, designed for induction cooktops, and heavy cast-iron pans had this problem.
It would be easy to turn the power on and off more frequently than once every 30 seconds. It wouldn’t be much more difficult to have a mode that delivered less than full power.
A thick iron plate under the pot smoothed the power delivery to the pot, but then it’s not really induction heating of the pot: just a hot plate.


Pancakes, steaks and simmering anything were my biggest frustrations.


It’s good to know there are some decent productions. That’s what I was expecting. I understand the technology. I know what is possible. It was very frustrating and disappointing. The largest appliance manufacturer in the world is selling poorly designed rubbish. The documentation doesn’t describe essential parameters. If only there were a practical way to distinguish the good from the bad before buying… try before you buy is my advice.


I had a terrible experience with one just last year. Had to replace it. Went back to an electric cooktop with simmerstats just last month. We’re much happier now. Can cook again without all the burning and boiling over.
I know an induction cooktop could be much better but the one we had couldn’t simmer anything: it could only intermittently overheat it. And occasionally it would switch to either full power (very dangerous - it was very powerful) or no power (absolutely ruins a steak when you’re trying to sear it). Technicians came multiple times and concluded ‘there’s nothing wrong’. Fortunately, after almost two years, they agreed to an ‘upgrade’.
I expect commercial induction cooktops are much better than consumer grade but they’re too expensive.
I wouldn’t buy another that I hadn’t tried first. I know one place that has a showroom with everything powered. Not that they would let me actually try cooking anything, but at least one can put a pot of water on and try out the controls.


I haven’t used 13/trixie but with 12/bookworm I use jmtpfs but it’s also available for trixie.

Only if the companies / people developing it are allowed to continue stealing IP without compensation.
It leaves out the steps where all the users’ data is sold to data brokers and leaked to criminals.


Rescued Slaves Crowd the Deck of the HMS Daphne, 1868 explains the circumstances in a little more detail.

Most of the images in the tutorial don’t load: access denied. Makes it a bit hard to follow.


Thanks. I see that is in the linked article. My mistake was to read the article linked in the first paragraph: with the link text ‘restricted the approval’, thinking that was the details of the restriction.


According to the linked in the article, the restriction is a requirement to revise the warnings
WASHINGTON (AP) — The Food and Drug Administration said Wednesday it has expanded existing warnings on the two leading COVID-19 vaccines about a rare heart side effect mainly seen in young men.
In April, the FDA sent letters to both drugmakers asking them to update and expand the warnings to add more detail about the problem and to cover a larger group of patients.
I don’t see any indication that the vaccines are not approved other than that the accompanying documentation must be changed. The companies have now had several months to make the required changes.
The linked article suggests that the conduct of the FDA and resulting requirement to change the documentation was inappropriate. But there is no restriction other than the requirement to update the documentation. Or am I missing something in the article?
It’s an interesting article but it seems to me that when it comes to opposing abuse of power, free communication is more fundamental than free software. Without sufficiently free communication, free software is practically unavailable and for many purposes (anything that involves communication with others) it is unusable. Without sufficiently free means of communication, the fediverse will cease to exist. Access to and use of the Internet is increasingly regulated.