phil21

I helped implement the concept of a "Sunday Supper" with our friend group in the city. It's basically held every Sunday and rotates to different households who feel like hosting that day. Luckily one person is sort of the "default" since their family's child situation makes it more convenient for them to host vs. go elsewhere.

It's basically an open invite for a group of friends, and a standing thing that just happens. If you have no other plans, you basically go. Sometimes I have to drag myself out of the house to do it - but it's worth it every time. The group that attends sometimes is just one other couple, sometimes it's a house full of 30 people if everyone comes, brings their kids, and a few friends of friends tag along.

Definitely one of the better things I've been involved in!

> Especially when they cross into the physical realm as in not properly secured and air gapped control systems. SCADA is scary.

What about the bad actors (choose your own evildoer here) who purposefully do not air gap their agents? And specifically train them to attack in such a manner?

I'd much rather have relatively benign stuff like this hit first, because the former is coming sooner than later. It's already here in a limited manner, likely more than any of us currently realize.

Botnets could crack passwords faster than anyone thought possible over 20 years ago now. This is just the latest iteration of such a concept.

There is so much low hanging fruit in this space that frontier models are currently utterly irrelevant. It's going to take decades of human-speed securing of IT to make superintelligence or whatever you want to call it a necessary component for such attacks.

At this point, someone with a rack or three of GPUs with 100kw to burn can replicate such attacks if they feel like it. the bar for entry is not even 7 figures.

It’s almost always both. Basically every single retired person I know in my family or friends of family are relying on their home to be their primary asset/income in retirement. It’s always a much larger chunk of their net worth than any retirement savings, if they even have the latter.

They were commonly called forced savings accounts when I was growing up. The standard advice was to buy a home as an investment vehicle. Every older family member giving me advice would preach about how they don’t make any more land and the price will only go up.

You can - to this day - talk until your are blue in the face to these people that your primary place of living is not an investment, but they will refuse to even entertain the idea even with the stark cold math staring at them in the face.

If these retired folks can’t sell their homes for a huge profit they will be in for some very difficult times a decade or two from now. They are highly motivated to keep the music playing as long as they can vote and show up to local city council meetings.

Rehabilitation has never really been the point of the US prison or even correctional system at all.

The fact you can serve your time, earnestly work on rehabilitating yourself, jump through all the hoops, etc. but still be under the economic death sentence more or less summarizes how people feel about it.

It’s basically NIMBY applied to the justice system. Everyone loves to pretend they want rehabilitation but also wants someone else to do it. This shows up in hiring, apartment leasing, social benefits, etc.

> The U.S. locks up people at 5x the rate of South Korea. Whatever the difference is, it’s not because the U.S. is unwilling to punish people.

The difference is in consistency in enforcement and general background crime rate.

If you have a nearly guaranteed chance of getting caught doing something, the punishment almost doesn't matter. If every single time you went over the speed limit you got a $500 fine 100% guaranteed, there would be effectively zero speeders ever. If you have a 1 in a million chance of getting caught but the punishment is 10 years in prison - there will still be plenty of speeders all over the place.

Humans do not do well with estimating risk on low probability events. We are all just big children in the end. Consistency in consequences are what actually matters, not overly harsh punishment.

There was/is a loophole where if you signed up via the iOS or Android app, it allowed it.

It's been disabled for some time now though otherwise, I check about once a day myself and keep and eye out on social media.

Annoying since I was about to upgrade back to the $200 plan after downgrading to the $100 plan due to being on leave and not needing as much usage the month prior. Doh.

> For all my efforts to get LLMs to work for me, they've failed. I figure it is because the tech space I work in (Industrial Automation) simply doesn't have the training data that someone building webapps does.

Putting safety critical aside (given that is a giant aside) for the moment, I'm a little surprised at this.

One of the most useful things I've found for AI Agent stuff is having an agent live on my home network and finally get the hundred or so random sensors and other IoT devices finally configured and talking together.

It's definitely not perfect though, and not anything I would trust my life to. This year however, I was able to set up enough sensors in my garden to have by far the largest hot pepper harvest I've ever had by min/maxing watering and fertilizing via way too many sensors and data fusing. This was mostly describing the sensors involved, the outcome and dashboarding desired, and pointing it to the various vendor API docs.

I could have done all that stuff manually, but never would have gotten around to it since it was just a tedious boring task. But with AI help I now can tell you precisely the soil moisture level, EC reading, ambient temp, etc. for each of my planters. And with some combining of my rain gauge data and sunlight sensing, tie that into automated irrigation runs. Then the bot just pings me when it thinks the next fertilizing cycle is due.

I don't know if that's revolutionary, but it took a project that would take me weeks to accomplish myself due to around a hundred sensors to configure/setup/etc. to a couple nights of messing around learning what all this talk was over these AI agent things.

I think there are plenty of use-cases where if AI gets things wrong 5% of the time it really doesn't matter a whole lot. I have plenty of these areas at my job.

Sure, only if you haul away the other 14 giant objects as well! And at least lie to me well enough that they will all go to a good home :)

It’s such an annoying problem to solve.

I have all sorts of “stuff” I would give away to someone who can use it. I just don’t want to spend 3 hours per object doing so via posting on Facebook marketplace, having 8 people flake out on pickup, and a Saturday afternoon killed because I need to wait around for someone to show up.

It gets worse with more esoteric things like nerdy electronic gear which is absolutely worth something to the right person, but that person happening to be in my area and also searching for that particular bit and bob at that exact moment are almost nothing. Some of these things can be hundreds of dollars if you need to order new or off eBay.

I’d be fine with giving it away to a reseller if I could. Just very difficult to throw it away since it’s so wasteful.

The most fitting description of an owner of a small coffee shop is "barista". Most barista's I know also handle far more than just making coffee and serving customers. They do everything short of being personally financially responsible for the business itself.

It's quite clear that if anyone was operating in bad faith here, it's not the commenters challenging your position. Small business ownership is almost always simply a riskier version of having the job you're also hiring for with a very small chance of upside. Usually at lower pay.

In this particular case the owner was simply doing a task other barista's like herself were doing as part of the regular job of a barista.

> The owner was not personally either making or serving the coffee.

This is the part I do not follow. Of course she is? What small coffee shop owner do you know does not personally make or serve the coffee on a regular basis? The only way not to take your original post in bad faith is to not understand your meaning.

Funnily enough, I was thinking of exactly those photos when writing my comment. I never found them to be all that misrepresentative - that they took a photo from their family friend of their father taken 20 years ago for the same dish they are making me and have it on some faded menu board behind the counter is... honestly pretty close.

It's the fake AI generated "shine" I guess that is the difference to me. Obviously totally made up, not real food, and made to look pretty.

It's mostly the uncanny valley. Both are indeed false advertising, but the AI generated ones are so obviously made up and not real it's a little off putting to me.

Wouldn't really stop me from ordering if I were already there, but it might make me order from their competitor if I were on Uber Eats or whatnot.

I don't follow. Even a little bit.

Every single small coffee shop owner I know has employees. They work alongside them and do every job in the shop along with them. They operate in shifts, sometimes the owner is there for both open and close.

What does hiring a barista have to do with not making coffee yourself?

Gonna go with the "hilariously out of touch" comment. These places rarely print money to the point an owner can be an absentee manager type.

I think there is a big difference between using AI generated food pictures, vs. taking real pictures of your food and using AI to create a menu/poster.

The former is both weird, creepy most of the time, and false advertising. The latter is as you say - the state of the current status quo. I can tell that it's AI, but honestly it's just replacing low-effort human crud for low-effort AI crud. Doesn't bother me at all, it's actually a great use for the technology imo.

> The utilization rates of GPUs for AI workloads in a datacenter run by cloud service providers (CSP) is between 60% and 70%. With such utilization rates, a GPU will typically survive between one and two years, three years at the most, according to a quote allegedly made by a principal generative AI architect from Alphabet and reported by @techfund, a long-term tech investor with good sources.

This is just extremely unbelievable to me. I am certainly not operating at a level the hyperscalers are and have much more limited direct experience. But I do actually put various GPUs inside datacenters (and much harsher locations) and have operated them at balls-to-the-wall 100% utilization for over a decade now.

You get the typical bathtub curve of failures. Unless the hyperscalers are operating these things even more overclocked and beyond thermal specification limits than early GPU crypto miners used to do, I simply cannot believe that the average hardware life is less than the useful life of the whole chip generation itself.

I have plenty of decade old GPUs that operate today just fine. Both consumer and datacenter form factors. The failures tend to be board-level like capacitors and such, so if you are operating at a massive scale partnering with someone who can fix those relatively cheaply is not all that difficult either.

It could be that these H200 and above class sort of stuff is engineered extremely fragile, but I seriously doubt it. The prevailing "common knowledge" pre-AI for GPUs were that they'd burn out in a year or two of heavy use, and that was simply untrue. I saved a ton of money buying batches of used units because everyone was terrified of this - and had no more early failures than I did buying brand new after basic refurb of re-pasting and putting a new fan on them.

Only if you use dollar value output as some weird substitute for actually making things.

Mostly bug fixes relevant to higher end home theatre setups. Typically bleeding edge stuff where a TV is released before a new HDMI spec is fully baked or whatnot. One of my older OLED panels needed a number of updates to get proper Dolby Vision profile support a few years back.

Typically not all that relevant most of the time to the average consumer, but if you’re having a persistent annoyance it’s worth looking through release notes from time to time.

This ignores that the vast majority of consumers actually use the smart tv features like the Netflix app. Then updates start to become far more relevant and eventually mandatory.

It also gates some "features" like the "AI" auto-detection of media type to automatically switch the TV between sports, movie, gaming, etc. modes as well. At least from my memory of a recent LG OLED setup.

So they can use Netflix or HBO apps and whanot.

The vast majority of consumers are not going to spend another $100-300 on a media player device like an Apple TV, Roku, or nVidia shield. They will use the built in app functionality.

And even those who don't - a very surprising to me number people use the "free channels" most TVs ship with. It's like the streaming version of broadcast OTA I guess for a certain consumer segment.

It was a stupid regulation to begin with. Any investigative agency already could trivially get this data through already existing means.

All it did was make your 3 member condo association have to pay a lawyer $500/yr in compliance fees for utterly no benefit to anyone other than the compliance folks collecting their rent. Or take on personal criminal liability to file it yourself. Better not forget one random year or it’s a felony!

Just like most of the rest of the AML/KYC creep after 9/11 it did absolutely nothing to go after anybody they say it did - it just made things harder on normal everyday people.

All that stuff is more dirty to the local community than your average datacenter.

The vast majority of these these no one knows even exist until told. They are basically warehouses without loading docks not much traffic goes into or out of.

Sure some horrible projects exist and are repeated over and over on social media, but they are the exception.

Most folks working across the street from one of these facilities had no idea until they were told to care.

Depends on which video codec. If it’s the most efficient one that has dedicated silicone embedded in the chip, it’s going to work great.

Web browsing is typically significantly harder on power. Plus these benchmarks usually mean in a perfectly dark room with the backlight at close to zero.

Yeah you’re describing normies. They are the norm. It’s in the name. People as described make up the vast majority of the population.

Nerds have always been in the extreme minority. Ignoring the influx of cosplaying MBA financing bros and industry hangers-on as tech became the place where the money is of course. The tech industry started attracting the normies as soon as it became almost as lucrative as finance, so these days the definition gets a little blurry.

If OP means Ethernet over HDMI it’s in the spec but as far as I’m aware has not been implemented in any mainstream chipsets anything actually uses. It’s one of those theoretical things that is somewhat surprising didn’t happen.

Overall it’s not a bad idea if devices need network connectivity. Have one central “box” get an Ethernet drop and it acts as a switch for your tv/xbox/playstation/media player/etc. The spec was made back when A/V receivers were more popular than they are now. It would cut down on a cable per device.

If OP means an exploit/hack then that’s a bit different.

I see it as the first step towards removing entire workflows like this.

Once the humans are removed, there is far less motivated reasoning to keep that process alive if it’s just robots burning tokens passing paper back and forth.

I will likely somehow be proven wrong, but it is one of the main uses I am optimistic for in the AI space.

Wouldn’t be all that surprised if quality went up with the use of AI for this stuff.

The word “slop” is tossed around way too much, but these sorts of engagements are a perfect example of replacing low value human slop with computer generated slop.

Probably a net positive to society because a few interns at a consulting company got to spend a Friday night enjoying themselves vs. working on a BS report.

Yes, I might be slightly jaded. But I’m not certain I feel this is hyperbole either.

No, just a latent bug that got exposed by a coincidental condition which happened to be synchronized via time.

Doesn’t require and patching at all. These things lurk for years or decades before finally meeting the conditions needed to be exposed. Many times one-off failures that underwrote the issue are written off as hardware failure or whatnot and not actually diagnosed until something like this happens.

It’s not revealing at all because unemployment numbers are directional and only useful in terms of comparison to prior years. The actual headline number is arbitrary and pretty much any metric you want to track already exists but is less popular. The only thing that matters is the methodology stays the same over the long term.

It might be educational to some, but this has been political rage bait since I became politically aware enough to listen to news, talk radio, etc going back to the 90s. It’s always been a favorite talking point for certain types.

Switch to using U-6 if it makes you feel better about the headline number making sense. U-3 has always been very clear over what the number is tracking and the methodology is quite well understood.

Could also just be a bunch of IoT devices running on synchronized time source that have what amounts to a defrost cronjob. And a latent bug that due to everything being synced and on the same schedule failed in the same manner (eg crashed the controller immediately after turning the heating element on and either didn’t boot back up or booted and some shitty programming forgot to check the heating element status). Saving state across boots or having a race condition of some sort like this gets my vote - since it’s a pretty common failure mode junior programmers are susceptible to.

I’ve seen similar too many times in my career - synchronized clocks are great until someone deploys a cronjob that fires at exactly the same time across the fleet and it breaks a certain OS or firmware revision. Which then exposes another bug that exists globally.

I definitely still subscribe to don’t attribute to malice what you can attribute to incompetence mindset.

Certainly all speculation though, real root cause will be interesting if it ever becomes public.

Yeah. There were times in my life I would have put my life on the line (and that of the thief) to defend my vehicle and the tools within it. Those fed myself and my family. I would have had no reasonable means to replace my tools, and barely reasonable means to replace the damage a catalytic converter theft would have caused me.

Today it's trivially easy for me to say "property can be replaced, a life cannot" since I am exceedingly privileged. I would not put my life, or anyone else's on the line to defend anything but my and my family's personal safety. But I remember what it was like to not be so ridiculously privileged. So I get it.

Only someone who has never lived a hard day in their life can spew such stuff at other folks. It's entirely circumstantial.

The laws should reflect the prior case, not the privileged one. At least if you want to disincentivize criminal behavior.

The world (I guess politics, but more broadly how you envision society to fundamentally function) is far more broad and nuanced than the silly (American) liberal vs conservative buckets.

Just adding a “libertarian vs. authoritarian” axis to both labels changes things up wildly. And shows me at least a little bit how someone can label themselves conservative and also be totally fine with the surveillance state. Or totally against it.