cmcaleer

I'm sure an automated system would do great in 95% of conditions but it's that 5% where you really, really need a human in the loop to make judgement calls that would be incredibly difficult to program.

Here's a classic example of a controller noticing a pilot is hypoxic, indirectly testing his competency and ability, and likely being careful with how he routed traffic around the unreliable pilot until he got better. This alone seems pretty hard to imagine automated with current technology without some overkill prone-to-failure solution.

https://www.youtube.com/watch?v=GVpfOvVgHtY

It's as much a gun to adversaries' heads if they hold lots of your bonds (China used to hold over $1T in USTs) so they would have an interest in not having that blow up by not annoying the US so much they refuse to pay foreign bonds.

It's also kind of neat as a financial MAD instrument, because China obviously can't just get everyone to go into their brokerage and market sell a trillion in USTs so they get a lever to pressure the US with but they can't go nuclear without also vaporising their own value. So it works as a melting ice block on which diplomacy is forced on both sides.

The death of the UST as a reserve instrument would be a catastrophically bad loss, the only silver lining is that the melting ice block might still be enough by the end of this admin that the new admin might have enough to stand on to rebuild it should they realise the value of it.

I remember being blown away when I was in a Kyoto Familymart after a few months of living in Osaka after they handed me my fried chicken very delicately with both hands like it was a business card!

I guess that’s the cultural divide that occurs when one community is fishing and trading while the other does, like, competitive perfumed calligraphy or whatever.

I think if you were to do an Osaka version of this, the list would be limited to maybe 4 of these (licking, chopsticks upright in rice, passing between chopsticks, and pointing esp. toward a senior would be taboo).

Whereas when I had a date with a girl from Kyoto, one of the first things that happened when we went to eat was she had to stop me from picking up my chopsticks impolitely and show me the proper way of doing it.

Suffice it to say my Osaka-learned table manners and speech patterns meant there was no second date.

I think if, 10 years ago, you spun Microsoft into several different companies with everything playing out exactly as it has today in the product management side, the most direct consumer-facing sections like Windows Desktop and Xbox would have cratered and most analysts would say that they have bleak futures, while Azure and 365 would have grossly overperformed and would have been titans.

MS has been successful despite fucking up the monolithic position they held in desktop and gaming, because they managed to find a particularly valuable golden goose. It's just that in doing so they allowed the other golden geese they have to become quite sick.

If you took out cloud rev MS would have been much more motivated to not let the rest of the company's products turn in to the sorry state they're in.

Tesla has been this for a long time since the price fundamentally does not reflect the quality or future of the company, yet the road is littered with dead Tesla bears.

Efficient markets hypothesis breaks down with Musk companies.

Agree they’re great but it does fascinate me when there are weird edge cases that Apple mess up.

For example if I have my phone and laptop running, and I’m listening to something on my phone, I pause with my AirPods, and then I unpause with my AirPods, instead of what was playing on my phone resuming through my AirPods, a video that I’ll have forgotten about will instead play through my laptop speakers, and pressing pause on my AirPods will do nothing and I have to interrupt whatever I’m doing to pause on the laptop. Possible they’ve fixed this specific issue though since I’ve learned to not have anything that has media controls open on my laptop.

The cross platform control stuff is probably very hard and usually works though.

I gave up on USBC headphones because if your port becomes full of lint (say by being in your pocket all day), it doesn’t take much to disturb a USBC connection and cause it to go through the whole handshake all over again for a few seconds.

Compared to 3.5mm where the frustrations I remember were usually limited to sometimes getting a bit of a crackle or one of the audio channels dropping out and worst case scenario you just unplugged it and put it back in and it usually worked. With USBC you have to wait to see.

Gut cleanses are probably stupid but I wonder if people would benefit from taking antiparasitics prophylactically. It's not something I've ever done, but I eat sashimi pretty regularly and wonder if I should take something like praziquantel because I'm probably at risk for Japanese broad tapeworm, and the symptoms are mild enough I can't really tell without testing, but the price of actually testing is much higher than just taking a drug with a great safety profile.

For similar reasons, I also wonder about people who consume raw milk. These people are more likely to endorse ivermectin for e.g. covid, because it made them feel much better. Maybe it's possible these people aren't lying about that, but not because it cured their covid.

These days I use Apple hardware despite Apple’s software, not because of it.

Better yet, a link to Kaggle and provide prize funding for a few dozen competitions with most of them open to UK residents only. Directly incentivise the most driven types of people to compete and learn and give local firms a way to identify talent.

But I guess donating another £4MM to PwC is more sensible.

You could have contracted 5 small firms for £400k each (which, for this project seems frankly seems excessive) and even if a couple failed to deliver you'd have gotten 3 separate products to choose the best quality one from, £148k to legally chase up the firms who failed to deliver, and still had £2 million left over.

I agree a good solution isn't easy to come up with, but the status quo is certainly an outrageously awful one.

I'm struggling to follow the logic on this. Glocks are used in murders, Proton has been used to transmit serious threats, C has been used to program malware. All can be legitimate tools in professional settings where the users don't use it for illegal stuff. My Leatherman doesn't need to have a tipless blade so I don't stab people because I'm trusted to not stab people.

The only reason I don't use Grok professionally is that I've found it to not be as useful for my problems as other LLMs.

You’re moving the goalposts. Gary Marcus’ proposal was being able to ask: Who are the characters? What are their conflicts and motivations? etc.

Which is a relatively trivial task for a current LLM.

I don’t really see how this is an AI issue. We use AI all the time for code generation but if you put this on my desk with specific instructions to be light on review and it’s not a joke, I’m probably checking to see if you’re still on probation because that’s an attitude that’s incompatible with making good software.

People with this kind of attitude existed long before AI and will continue to exist.

The following fees apply when a user completes [...] any app installs within 24 hours of following an external content link

So does this mean a malicious competitor or motivated disgruntled user could fraudulently cause millions of app installs? With the scale smartphone activity fraud farms are at these days, paying a few thousand dollars on such a service to cause a developer to spend a few million dollars on worthless installs (or a lot of resources arguing with Google) seems like a worthwhile endeavour for the motivated.

The only thing that gives some slight semblance of hope is that he at least acknowledges that Mozilla is vulnerable and he very very briefly mentions needing new sources of revenue.

No mention of an endowment (like Wikipedia has) or concrete plans to spend money efficiently or in a worthwhile way, and I sure hope ‘invest in AI’ doesn’t mean ‘piss away 9 figures that could have set up an endowment to give Mozilla some actual resilience’.

I hope is that he’s at least paranoid enough about Mozilla’s revenue sources to do anything about their current position that gives them resiliency. Mozilla has for well over a decade now been in a pathetic state where if Google turns off the taps it is quite simply over. He talks a lot about peoples’ trust in Mozilla. I don’t really remember what he’s talking about to be honest, but if Mozilla get to a point where they seem like they can exist without them simply being Google’s monopoly defence insurance, perhaps I’ll remember the feeling of trusting Mozilla. I miss it.

Being short anything AI now seems like shotgun tasting unless you really want to give Citadel and Jane Street money since the options premiums are so high, but I have been trying to get a bit less exposed to tech over the last few months and just been buying other ETFs that are less exposed to tech.

MS (or any large company for that matter) didn’t participate in BLM discussions and get speakers to describe themselves and list their pronouns because they thought it was virtuous or right, they were just following the cultural zeitgeist in a way that they thought would make them more money.

Walking it back is just the same behaviour manifesting in a different way. Investors don’t value DEI in the same way they did before so it becomes an expense with no value to shareholders, so it gets cut.

It’s very cynical but nothing about this should be particularly shocking.

You don’t need to loudly and publicly say no to Roskomnadzor’s extrajudicial notices, by recognising them you’re giving them more legitimacy than they deserve by acknowledging at all rather than treating the notice as the spam it is.

Just because the UK is modelling themselves in the same image doesn’t mean that the tactic for dealing with extrajudicial censorship attempts is different.

I moved away from Anker to UGREEN following Anker’s Eufy using unencrypted feeds and sending data to the cloud with no user consent, which was bad, but their gaslighting response to the tech media and overall handling of the situation made me completely lose all trust in them. Maybe they’ve gotten better since, happy to be proven wrong.

UGREEN fit that niche of ‘tech products that are generally of quite good quality’ for me fine. They feel like neither an upgrade or a downgrade to the Anker stuff I still have.

I shitposted about the ‘year of the Linux desktop’ for most of my teenage and adult life only to dual boot Linux for a bit as an experiment a year or two ago and have ~never booted in to Windows for anything other than games with anticheat since.

It’s more a case of Windows getting significantly worse and feeling like malware than Linux desktop being THAT much better than the last time I tried it, but between proton and maybe 95%+ of my work being in a browser window or browser window (Electron) these days, I basically never run in to compatibility issues and never have Candy Crush advertised to me.

OP has given all that they need, English is probably not their first language, and the first response was to ask for even more unpaid time and labour (presumably in English) in a format that is likely much more difficult for them to summarise coherent thoughts succinctly in, while addressing zero of their issues. They’re asking for MORE effort of the volunteer to fix their bot’s fuckup.

I’m honestly struggling to think of a more insulting way to respond to this. At least “Fuck off” isn’t pretending to care, it’s fewer words to read and isn’t asking for an indeterminate amount of time from you.

What’s stopping a similar crisis that 23andMe customers faced where their genetic data along with their identifying information getting sold to the highest bidder if you ever become insolvent?

Considering how big a deal that was at the time, and how strong a differentiator it would be, it’s notable how absent from the homepage. It’s nice that Nucleus claims not to sell the data, but 23andMe had similar claims, it wasn’t strong enough to prevent genetic data from being transferred if they were to be acquired.

This has always been something I’ve been interested in but so far no company handles privacy concerns of data that’s so deeply fundamentally personal and private in a satisfactory way, and I’m especially apprehensive post-23andMe.

Taking a dice roll on Nucleus not just pulling another 23andMe seems not worth the ~$3000 saving you claim to be offering.

It is frankly ridiculous how unintuitive it was to add an email account to Mail on iOS. This is possibly the most basic functionality I would expect an email client to have. One would expect that they go to their list of mailboxes and add a new account.

No. You exit the mail app -> Go to settings -> apps -> scroll through a massive list (that you usually just use for notification settings btw) to go to mail -> mail accounts -> add new account.

Just a simple six-step process after you’ve already hunted for it in the mail app.

Open Office still exists (and is being actively supported)

Actively supported is a stretch. Look at these commits.

https://github.com/apache/openoffice/commits/trunk/

This sounds pretty helpful. If I'm trying a new lib that I want to do something specific with, I paste all the documentation in and interrogate the LLM about it, then cross-reference with the docs. Usually much faster to do what I want than just CTRL+F or writing a SO question that gets immediately marked as duplicate because some other question is vaguely related.

For language learning, it's terrible and will try to teach me wrong things if it's unguided. But pasting e.g. a lesson transcript that I just finished, then asking for exercises based on it helps solidify what I learned if the material doesn't come with drills.

I think writing is one of the things it's kind of terrible at. It's often way too verbose and has a particular 'voice' that I think leaves a bad taste in peoples' mouths. At least this issue has given me the confidence to finally just send single sentence emails so people know I don't use LLMs for this.

My frustrations with LLMs from years ago has largely chilled out as I've gotten better at using them and understanding that they aren't people who I can trust to give solid advice. If you're careful about what you put in and careful about what you take out you can get decent value.

I always thought that Fight Club quote missed a crucial extra number: D, the cost of brand reputational damage. Maybe best expressed as multiplier <1 on revenue that converges towards 1 over time.

Maybe a decade from now this becomes a semi-obscure bit of gun lore and SIG US more or less recovers.

But right now?

If you’re in the market for a handgun, or any gun for that matter, are you going to touch SIG US’s stuff? Maybe, but there are customers who might just say that they don’t want to take that chance and go with Glock or whoever else. Do those customers come back? Why doesn’t this get brought up at every single bid for a contract SIG US makes from now on?

It’s possible that they’re still making the right EV call of course, but the medium-term hit they’re taking from a flaw like this that you don’t make right has to be massive.

Here's a question on expected value. Do you think Epic makes more money if:

   (a) they agree to Apple's demands and have Fortnite on the App Store during its peak of popularity for years and eat the junk fees on mtx or 

   (b) they fight an extremely costly lawsuit, which they have no guarantee of winning, for years, during which time Fortnite could leave the cultural zeitgeist (which it to some extent has) and maybe eventually one day get closer to 95% of mtx money?

Monetary damages are damages, I don't think this is particularly complicated. If I made it so you couldn't get several weeks of your wages for hours that you worked you would be rightly furious with me and feel like a victim.