> The interesting part: our air-gapped deployment actually runs faster than our SaaS version.
This is the least surprising thing I’ve read all day.
1984: U.S. withdraws. 2003: U.S. rejoins. 2011: U.S. stops paying dues after Palestine joins. 2017: U.S. announces withdrawal (effective end of 2018). 2023: U.S. rejoins, pledges to repay dues. 2025: U.S announces withdrawal
Seems to be a revolving door
The US is complicit in the intentional starvation of gaza’s people by israel. At least 15 people have starved to death in the last 24 hours, including an infant.
https://www.cbc.ca/news/world/israel-gaza-war-hunger-childre...
Ragequitting UNESCO over their recognition of palestine is a small part of the project of supporting the ethnic cleansing of gaza and the west bank.
This administration has zero ability to build or bring people together, able only to destroy what others have made. On the bright side, it isn't effective until December 31, 2026 so there is plenty of time to chicken out.
I am surprised how many people have so many problems with Firefox.
I've never felt impeded by loading speeds, and my ADHD regularly has me forgetting to restart it, to the tune of 100+ tabs open across multiple desktops. My wimply little MacBook Pro doesn't seem to mind.
The only downside I've found is that, because so many people just default to "Chrome or nothing," there's occasionally sites that have bugs because, like was the case in the 90s with Internet Explorer, the site developers took the idiomatic Chrome way of building a feature instead of something universal.
The same happens with whisper-large-v3 on Chinese transcription: silence is transcribed to something like "please upvote, share and favourite this video". I suspect they trained the model on some random YouTube video without carefully picking really useful data.
I work in manufacturing in the US. Incoming quality control, for Chinese vendors, is necessarily set up with zero-trust. This isn't a "trust but verify" sort of thing, it's strictly "do not trust". Assume that, at every step of the chain, there will be a lie: change of process, material change, collected data, and that the product being given to you is even yours (delivering a knockoff at the final step, and reselling yours on the gray market).
This is all common knowledge, proven by example after example that it's necessary to have zero trust. It's truly an adversarial system. All the extra engineering effort for IQC is still cheaper. And, there's rarely an alternative to the amazing manufacturing ecosystem that is China.
After tracking down several of these types of issues, it appears that the Chabuduo mindset [1] is a very real thing.
[1] https://news.ycombinator.com/item?id=32465780
FWIW this has caused a big storm in China. The root of the issue is known to be caused by the battery cell vendor Amprius changing the battery design w/o notifying the power bank manufacturers. AFAIKT Amprius lost the 3C certification (a certification in China) because of this incident.
This is one of the Chinese reports on the issue: https://m.thepaper.cn/newsDetail_forward_31048287
Excerpt from the report above (translated using Google):
> The Paper learned from an insider that Anker Innovations' battery cell supplier is already a leading battery cell supplier in the industry, and did not inform customers after it changed materials. In addition to Anker Innovations, the supplier also cooperates with leading power bank brands, so the impact is huge. Although Anker Innovations did not name the supplier, an insider pointed out that the supplier was Amprius.
UPDATE:
There's an exclusive interview by 36kr with one of Anker's VPs:
https://m.36kr.com/p/3365435892680709
Sri Lankan here. I used to compete in the IOI (International Olympiad in Informatics), back in the '90s.
I aced the competition in Sri Lanka, but I failed to win a medal the first time I competed internationally. I solved several problems with recursion, but these failed to complete within the strict time limits allowed.
One of the contestants from another country told me: "You needed dynamic programming to solve problems 1 and 5." I then spent the next year trying to figure out what dynamic programming was. The folks over here weren't familiar with the term. In fact, I was often asked "did you mean dynamic memory allocation?"
After a while, I managed to find a book that talked about dynamic programming, and I was like "Oh, it's recursion + storing results" and "Ah, you can also do this iteratively in certain circumstances"
Bottom line, armed with this new knowledge, I ended up winning a gold medal at IOI 2001. Fun times.
> With fewer visitors, Stack Overflow is seeing fewer questions posted on its message boards
When you operate a community that's hostile to questions that have already been answered, are poorly researched, or are homework, don't be surprised when people start taking those questions elsewhere, and don't be surprised when they start asking their good questions elsewhere, too.
For whatever reason, evolution decided those wavelengths should be overlapping. For example, M cones are most sensitive to 535 nm light, while L cones are most sensitive to 560 nm light. But M cones are still stimulated quite a lot by 560 nm light—around 80% of maximum.
The reason is simple: genes coding the long wave opsins (light-sensitive proteins) in these cones have diverged from copies of the same original gene. The evolution of this is very interesting.
Mammals in general have only two types of cones: presumably they lost full color vision in the age of dinosaurs since they were primarily small nocturnal animals or lived in habitats with very limited light (subterranean, piles of leaves, etc.) Primates are the notable exception, and have evolved the third type of cone, enabling trichromatic color vision, as a result of their fruitarian specialization and co-evolution with the tropical fruit trees (same as birds, actually).
So, what's interesting is that New World and Old World primates evolved this cone independently. In Old World primates the third cone resulted from a gene duplication event on the X chromosome, giving rise to two distinct (but pretty similar) opsin genes, with sensitivity peaks at very close wavelengths. As a note, because these genes sit on the X chromosome, colorblindness (defects in one or both of these genes) is much more likely to happen in males.
New World primates have a single polymorphic opsin gene on the X chromosome, with different alleles coding for different sensitivities. So, only some (heterozygous) females in these species typically have full trichromatic vision, while males and the unlucky homozygous females remain dichromatic.
Decent wikipedia article on the subject: https://en.wikipedia.org/wiki/Evolution_of_color_vision_in_p...
Types of opsins in vertebrates: https://en.wikipedia.org/wiki/Vertebrate_visual_opsin
Many comments here are arguing that quality has actually gone up over the past decades. However, a common experience for me is that I own something of good quality from 5/10/15 years ago and now buy the successor model from the same brand, but the product has gotten worse, being cheaper made. And I have a hard time finding a replacement that matches the quality of the old version. It’s a regularly reoccurring frustration.
My suspicion is that when products are successful and mature but reach market saturation, profit growth pressure leads to cutting some corners on every iteration, and hence to a slow decline in quality over the years.
An unreliable computer treated as a god by a pre-information-age society sounds like a Star Trek episode.
https://threadreaderapp.com/thread/1945959030851035223.html
The leverage is that the activists will potentially be able to draw the ire of the government. Visa and MC get away with absolute murder in terms of the size of the fees that they charge in the US. Most developed countries don't allow that. The US government could easily regulate them (as they already do with debit card fees) or use anti-trust law against the obvious duopoly charging exorbitant prices. Because of this situation, Visa and MC have a very strong incentive to crack down on things the government doesn't like.
The unspoken arrangement is that the government allows them to keep charging a de facto sales tax on a massive portion of the economy as long as they cooperate and de facto ban things that the government wants banned but can't ban themselves due to that pesky constitution.
Look. Ignore the content. Why the fuck do we allow credit card companies have a say in how we spend our money?
Fraud? Abuse? Fine, let me put cash onto a card and if that card gets stolen, oh well, my loss. Mastercard should have no say in what what speech is considered acceptable outside of their offices. We don't care what execs at a water company think? Why do we care about the people at Mastercard?
One factor is the ongoing campaigns from number of moral crusading groups who lobby them to cut off payment processing for things they don't approve of. NCOSE has been working for decades on the project, and targeting credit card companies has been a successful tactic for them for a decade or so.
[1] https://www.eff.org/deeplinks/2020/12/visa-and-mastercard-ar...
[2] https://www.newsweek.com/why-visa-mastercard-being-blamed-on...
[3] https://scholarworks.iu.edu/dspace/bitstreams/761eb6c3-9377-...
Why do payment processors do stuff like this? Is there some regulation that requires them to? I get that they don't want to process fraudulent transactions, but I'd think the response to a higher percentage of fraud from some industry would be to charge them more. It doesn't make sense to me why they would be concerned about the content of games, as long as everything is legal and the parties concerned aren't subject to sanctions.
Some of these games seem completely abhorrent, and probably illegal in more restrictive jurisdictions, but not the United States. And I've not seen any suggestion they're funding terrorism or something. So I'm perplexed.
> The amount of cognitive overhead in this deceptively simple log is several levels deep: you have to first stop to type logger.info (or is it logging.info? I use both loguru and logger depending on the codebase and end up always getting the two confused.) Then, the parentheses, the f-string itself, and then the variables in brackets. Now, was it your_variable or your_variable_with_edits from five lines up? And what’s the syntax for accessing a subset of df.head again?
What you're describing is called: programming. This can't be serious. What about the cognitive overhead of writing a for loop? You have to remember what's in the array you're iterating over, how that array interacts with maybe other parts of the code base, and oh man, what about those pesky indices! Does it start at 0 or 1? I can't take it! AI save me!
> One user, who asked not to be identified, said it has been impossible to advance his project since the usage limits came into effect.
Vibe limit reached. Gotta start doing some thinking.
The "spreadsheet" example video is kind of funny: guy talks about how it normally takes him 4 to 8 hours to put together complicated, data-heavy reports. Now he fires off an agent request, goes to walk his dog, and comes back to a downloadable spreadsheet of dense data, which he pulls up and says "I think it got 98% of the information correct... I just needed to copy / paste a few things. If it can do 90 - 95% of the time consuming work, that will save you a ton of time"
It feels like either finding that 2% that's off (or dealing with 2% error) will be the time consuming part in a lot of cases. I mean, this is nothing new with LLMs, but as these use cases encourage users to input more complex tasks, that are more integrated with our personal data (and at times money, as hinted at by all the "do task X and buy me Y" examples), "almost right" seems like it has the potential to cause a lot of headaches. Especially when the 2% error is subtle and buried in step 3 of 46 of some complex agentic flow.
My bank uses a fraud detection system that calls you if suspicious activity is detected on your account. It then asks you to call back a number to verify the account activity. Every time they call, they provide a different callback number. Searching for the callback number online yields only one result, which is the fraud detection systems web page telling you to NOT trust phone calls of any kind (their advice is solid, but it tells you to not respond to their own legitimate calls)!
I work for a home builder. Our single biggest problem is finding and hiring quality contractors that have enough skilled tradesmen.
There are dozens of electrical contractors in my area. But only two that perform work to our standards.
There is only one HVAC company that can meet our standards. Same for all of the other skilled trades.
Our framing crew is the best within a 75 mile radius. Other builders are constantly trying to poach them from us. We keep throwing money at them to prevent them from going to another builder.
Non skilled labor like landscaping and pest control are a dime a dozen. I just fired our main pesticide and herbicide contractor today because they couldn’t get it together.
Of course I had them replaced before I fired them but I had almost 20 options to choose from.
Unfortunately I can’t say the same about all of the skilled contractors.
Ah, neat.
If I'm reading this right, then the real big benefit of these things would be solid state magnetic storage.
The benefit of these things is they don't create a magnetic field while they do respond to magnetic fields. That means you can pretty tightly pack these things together without concern that they'll interact with each other. A light electric pulse could determine if the bit is a 1 or a zero and a strong pulse would flip the bit.
I'm guessing that due to this nature, these things would actually have pretty long shelf lives and near infinite read/write cycles since you are, effectively, just flipping atoms around and not actually breaking structures or dumping in charge.
These should mostly work with regular silicon manufacturing. The tricky part will be how tightly you can pack these things together before the reading structures start interfering with each other.