neepi

To be fair people are pretty damn unintelligent when it comes to verifying information. Despite my academic background I catch myself doing it all the time as well.

However LLMs amplify this damage by sounding authoritative on everything and even worse being promoted as authoritative problem solvers for all domains with a small disclaimer. This double think is unacceptable.

But if they made the disclaimer bigger then the AI market would collapse in about an hour much like people’s usage does when they don’t verify something and get shot by someone actually authoritative. This has happened at work a couple of times and caused some fairly high profile problems. Many people refuse to use it now.

What we have is bullshit generator propped up by avoiding speaking the truth because the truth compromises the promoted utility. Classic bubble.

Have you considered that quality tends to cost the same everywhere as it is fairly rare? It’s not commoditised.

The problem is crap people are cheaper in India than crap people elsewhere. And that looks good on the balance sheet.

And as we’re about maximising shareholder value these days then that’s fine apparently.

Amazon pricing isn't always that good and they lie about the discounts and retail prices, at least here in the UK. My US colleagues tell me the same is true in the US. I've actually found that general high street crap seller Argos here tends to have better retail prices than Amazon. I can just amble on down the road and get what I needed same day pick up in person rather than wait for a delivery to turn up.

Is this a regional thing? I don't really hear phones ringing in the UK any more. They're almost always on silent by default.

Probably AWS issue. Our production RDS postgres cluster did that a few months back. Just stopped responding on the network. Health checks were fine. AWS support was mostly useless and couldn't work it out in an hour, despite having their top tier enterprise support, so with customers down we had to create a whole new cluster and do a restore from backup which took 4 hours.

RDS is now gone. It's on a couple of EC2 instances with replication, hourly EBS snapshots and daily shipping to S3.

I'm loathed to use AWS's "encapsulated" services for anything since.

The point is if you're outsourcing to India you want low cost and are not concerned with quality. That's exactly why you go there. And India has plenty of that to give. No one there wants to work for those outsourcing chop shops, but it's an easy way into the industry so staff turnover is high as people move up the chain. So if you're outsourcing for cost cutting reasons, which they are, it's a constant cycle of low quality staff. That is the risk. Nothing to do with race or xenophobia. You put low costs above risk provision.

We can all pick and choose good stories. What about Tata Steel's total mismanagement of Tata Steel Europe? And JLR isn't exactly in good shape as you say it is. You just picked some numbers that sound good. And wrapped it in a xenophobia straw man.

Their banking op is a fucking mess as well. Had no end of problems with their card services which were rebranded HSBC.

Err they were breached most likely through Tata Consultancy's helpdesk apparently which is literally the people they outsourced it to.

Their approach was to sell the UK operation to Tata in 2018 and piss everyone off until they leave and replace them with Indian staff to save costs over time.

You get what you pay for. They're now paying for it.

You have to convert to an archival format from whatever your raw format is. The issue with raw files is that they are compressed so any bitrot in those leads to more than just losing a pixel or two here or there but sometimes the entire file. TIFF is not compressed at all. FEC allows single pixel corrections to be done accurately. We had a metric shit ton of LTO tapes, changers and online disk cache that handled all that.

I don't do that at home. I have 600GB of Nikon raws. I keep one copy on my Mac. I have a time machine backup in the house (in another room) on an Samsung T7 shield and I have an off site backup which is a quarterly rsync job to another Samsung T7 shield.

Yeah I'm not even talking about that side of things. I'm talking about after it leaves the camera. When someone shows you something, you should be able to say "I want to see the closest to original image". Humans make a lot of changes to images evidently to change perception. Even a simple crop can change the story being told.

A good example: https://imgur.com/iqtFHHg

Used to work on archival. You want to store your images in TIFF and use FEC if you really care. That withstands significant bitrot. Signing will just tell you that the image is broken but not allow you to recover it. You can do fine with SHA256 and 3-2-1 backup.

About the same. The 120Hz on the pro feels very slightly better and has slightly lower latency. This does help a tiny bit for detail work. But you get used to not having it after about 30 mins.

Yeah. Cintiq isn’t much use when you’re sitting in a field in the middle of nowhere sketching.

Yeah happy customer here for that combo. There is nothing that even gets close to it.

I await cryptographically signed photos coming out of cameras. Actually I’ve said that should be the case for the last 20 years.

You should be able to follow a chain of evidence towards an unprocessed raw image potentially.

I would do that but I dislike Cloudflare because they wanted by DNS as well. I keep my DNS / CDN separate. Too many eggs in one basket otherwise.

Well that's pretty funny, but is it really?

Most people within my circles don't use cloud storage any more, me included. This is a combination of being cheapskates, being fucked over on billing issues with no support (Microsoft), getting their account hacked (Google), actual tangible data loss (iCloud) or having genuine privacy concerns (all of them).

We all have USB sticks. Some of us use rsync :)

AWS free tier. S3+cloudfront has cost me $0.00 for the last year. This is incidentally the best price.

My (single page) personal site is HTML+CSS (no JS) based on a template generated by ChatGPT because I don't give a crap. Trying to make something that works on a mobile device and desktop is beyond my meagre skills. This worked fine.

I find it quite amusing. A couple of years back a good friend of mine went to Everest basecamp. That wasn't even the summit. So they have WiFi up there now and I get a message from her. The message was "made it" with a selfie attached where she actually looked like she hated her entire existence and wondered what the hell she was doing with her life. It was visibly rammed and apparently it smelled of human shit.

Expensive way to have a bad time.

No we dug out a faculty statistician and did the analysis and paid them even more money to fuck it off.

The consultancy resisted this horribly because the tech lead is a performative bullshitter and it doesn’t look good on his portfolio page that it looks like something from 2001.

Well the problem was that they got a new UI anyway when they didn’t need or want one.

I think that was the problem.

The mainframe terminal user interface was shown to me in all its glory by Italian gate staff at the airport in Rome. Literally hammered through a security check after I was flagged as having two return flights. It was no less than a form of kata.

I can imagine how shit it would have been if you have to log into windows and open a web app and use the mouse and stuff to click through a web form hacked up to do the job.

This reminds me of something that happened to us recently. So one of the internal things we have is a giant turd that sits on top of some Oracle/SAP mess. The user interface is circa 2001 and hasn’t changed really. Looks like a cross between Motif and GTK1 it’s that bad. The icons are off the shelf ones from some clip art collection.

So they need to rewrite it and got a consultancy to do it. The whole UI got rewritten using react and used an off the shelf flat design. A week after launch they sent out a user survey and nearly every respondent complained about the flat UI. So they made it look like the old one again very quickly.

So now we have a shiny new React app that looks like something from 2001.

I have to wonder if anyone likes flat user interfaces or just user studies are broken.

Supposed to be the same thing in the UK but no one cares. In fact most of our students (undergrad mathematics) appear to have HP Prime now which has a full CAS built in. The questions are designed to break the CAS sometimes. Try expanding (a-2b)^1000 on a calculator to get a coefficient out. It gets stuck and hoses the whole calculator.

Don’t count your chickens.

One of my former colleagues had the same thing. When he hit 52 he found out he had prostate cancer. After treatment he could no longer work. After a year of treatment he couldn’t afford health insurance or further treatment.

Fortunately you can buy firearms and shoot yourself in the face. Cost his wife and kids less that way. That’s the American dream.

That would assume that the entire world hasn't already tried this a thousand times over.

Measurements are standardised communication tools. If you start changing the definitions, things fall out of the sky and on your head.

No. dB is not a unit. If it's quoted as a unit, then the person using it doesn't know what they are doing. dBx where x is some reference value is a unit and that means dB relative to the reference unit. So dBm/dBv etc are units. dB is not.

Mathematically, dB is the ratio between two unit values and for example if you divide metres by metres you cancel the units.

Sorry but no.

Absolutely no one cares about the mathematical purity of a unit like the decibel. It's a hammer that turns log stuff into easy to work with linear stuff. It's x10 because that turns into reasonably easy to use numbers with enough precision to use casually. Otherwise it's just a log ratio. Any suffix is a log ratio to a reference value for the entire system (dBv / dBm etc).

Say in a 50 ohm system I have an level 7 mixer then I need 7dBm into the port. My oscillator kicks out 0dBm (1mW). Then I need a 7dB gain block. I go to the datasheet for the MMIC and pick out a bias network that gives me around 7dB. I grab a eval board for the MMIC, solder the resistors on it, shove this on my VNA and throw 0dBm in one end and see if 7dBm comes out of the other end at the frequency I need.

It's like Laplace. Do you really want to spend hours doing calculus?

Source: actual electrical engineer, not some clueless hobby blogger. I really wish people would stop writing authoritative sounding articles on stuff they clearly have little experience with.