reliefcrew

There was no clear pop-up, no "Are you sure you want to

You haven't learned MSFT's m.o. yet... this kind of pop-up only shows up when you're absolutely certain about something, like removing files. In those cases they ask for confirmation endlessly and needlessly.

When they're trying to collect money, they go ahead and quickly guess with all the defaults as they see fit. I'm surprised the default isn't 1000 seats tbh. After all, you're certain to be a huge success now that you've chosen them as a vendor.

;-)

Yes, they have data many don't have access to. And, they make "educated guesses" based on this data as well as the expertise available to them.

Their predictions may certainly be embellished as well. However, their statements can also be taken out of context or misinterpreted.

None of this is in question. The challenge is to make a careful judgment of your own.

Some things never change :-)

Say hypothetically the productivity boost from AI is 2x per employee

I believe this is a bad assumption. Most likely the boost is not normally distributed. So, although all employees become more productive, some employees become much more productive thereby making the others redundant.

As someone pointed out already... there are diminishing returns. In addition, there are marginal returns. Put the two together and you get ever diminishing marginal returns from the laggard employees.

A carefully audited risk management system should always sit in between trade generation and execution. So, direct broker access is definitely a terrible idea.

Vibe coding the generation of simulated trades is another matter. That has value.

https://mitpress.mit.edu/9780262560993/the-little-schemer/

She can play along on the computer by installing mit-scheme (http://www.gnu.org/software/mit-scheme/) and placing the snippet below into ~/.scheme.init

   (define atom?
     (lambda (x)
       (and (not (pair? x)) (not (null? x)))))

   (define lat?
     (lambda (l)
       (cond
        ((null? l) #t)
        ((atom? (car l)) (lat? (cdr l)))
        (else #f))))

   (define member?
     (lambda (x l)
       (cond
        ((null? l) #f)
        ((eq? (car l) x) #t)
        (else (member? x (cdr l))))))

   (define mber?
     (lambda (a lat)
       (cond
        ((null? lat) #f)
        (else (or (eq? (car lat) a)
                  (mber? a (cdr lat)))))))

Oh, yes... I didn't mean to imply that the newer "AI modes" aren't better and that we're somehow standing still. More that the differentiation is made much starker due to so-called "enshittification". Whether that contrast is deliberate or coincidental I'm not completely sure, but I'd wager it's the former.

It seems the relentless pursuit of greater profit has diluted "the old search interface" toward simply regurgitating adverts.

Customers realize this, even if subconsciously, and so the new "AI Modes" offered by search companies are required to stay relevant.

It's the process of creative destruction at work... round and round we go.

And if they fix it? Pay them. Simple.

Stop fooling around with abstract puzzles

I beg to differ on the simplicity here.

The very best candidates don't want to think about your stuff at all, unless they're paid ahead of time, since they've already made tons of $$$. In effect, they're already getting paid by your competition or they themselves are your competition.

It's a great example of Akerlof's "market for lemons".

The "satisficing" solution IMHO is to take [calculated] risk on available consultants/contractors and keep the better ones on retainer [to hire]. Yes, in a sense, I'm suggesting you pay [more] to interview people. That way you stand a chance of beating the average.

Finding the "best value" in a candidate is a trap... you only need good enough.

If the sender eventually proves to be a freeloader, they get added to an auto-respond list. I hope to enhance the auto responder with AI soon but for now it just tells them I'm busy and to bugger off. You never miss anything that important.

Is there anything I can do?

Have you tried reading Camus' essay "The Myth of Sisyphus"?

It certainly depends on the kind of writing too. Very technical writing should benefit from a computer. Poetry... probably not as much.

Obviously I can't speak for op, but... spellcheck and copy/paste alone make paper more difficult to work with.

Basically, writing involves a lot of re-writing and revision. Electronic documents make that easier.

The main practical win is that cmd args show up in `ps aux` for anyone on the system to see, whereas stdin keeps it off that list.

For those interested, re-mounting /proc with hidepid can prevent this:

    `mount -o remount,rw,hidepid=2 /proc`

I don't know how sleep/wake works from a text mode TTY machine if that's what you mean.

Usually, I expect, by just issuing a command... e.g. `systemctl suspend`

Old thinkpads are usually pretty good.

If you're looking at a chromebook I suggest doing research on the specific model first. Esp. by looking here... https://docs.mrchromebox.tech/docs/supported-devices.html

Old MS Surface tablets are also decent options IMHO... https://github.com/linux-surface/linux-surface/wiki/Supporte...

I'm presently writing on a Surface 3 w/ x7-Z8700 atom that's running Ubuntu Noble and vanilla 6.8 kernel. They're about $50. Everything works rather well (including gfx/audio/sleep/hibernate), except the camera, which I haven't bothered with. I'm quite happy w/ it tbh.

One thing to note w/ this model is you'll need the OEM keyboard to install linux since there's only one usb port; which will be required by the install media. Once you've installed though you can use any bluetooth or usb keyboard you want. Another thing is any micro-usb charger should work but I wouldn't rely on just 1amp, go w/ a 3amp.

Anyway, happy hunting!

what got you interested in having a mobile setup

Over time the need to quickly address administrative and operational issues while on the road converged with marathon coding sessions. So, I'd say it was organic and more "evolutionary" than "revolutionary". More often than not, I think advances tend to happen in that fashion.

Perhaps that's something to consider while you brainstorm. IOW, it may be good to compromise between a more grandiose vision and a mundane, but tangible delivery... then iterate.

Is this hedging bets a little? Sure, but I've grown into a somewhat risk averse developer so I like playing "small ball". Whether or not that's the best approach, I can't say.

I'll also "Ssh + GNU Screen + emacs" occasionally on my phone and I'll be honest... it's way better to throw a bt/mini keyboard in the mix.

After that you can have a mostly normal desktop experience w/ the addition of a tablet. How long can a person stare at a screen that's only six inches?

So, ultimately I'd suggest using a tablet and [mini]keyboard with the phone as a hotspot instead; unless you're a die-hard masochist. Of course it's a slippery slope and a full laptop might be ideal for most users.

Phone displays are just too small.

I've been wanting to build something in this space.

Did you have something specific in mind?

Interestingly, this: https://en.wikipedia.org/wiki/Apple_T2#Security_vulnerabilit...

reads as though it is possible using "checkra1n". But again... it might be easier to just buy something else :-/

I find "impossible" hard to believe. Perhaps it's so difficult that it's not worth the opportunity cost though.

You may have to remove the disk or use a JTAG programmer, or heck... maybe even re-solder something.

The more nuanced question in these situations isn't "is it possible" but "is it economically viable".

If that doesn't work either then it's a paperweight and you can sell it for parts

C'mon! It can't be that dire... at worst you should be able to reinstall or reset to factory defaults somehow. Granted the data may be lost, but the whole is still more than the sum of its parts :-)

I think "control" is a far too fatalistic description. Can people be manipulated, of course. Some more easily than others, sure. But... "astra inclinant, sed non obligant"

One of my favorite Show HN: HN dark mode in 2 lines of code...

https://news.ycombinator.com/item?id=32719488

It also solves the "Why no light mode?" problem ;-)

Just ask AI how to mirror w/ wget. But, beware that if the site relies on javascript, wget may not be enough. In that case you'll need to program some kind of headless browsing. Didn't the internet archive (archive.org) take care of everything for you already though?

Don't forget lying!

Cool, ffmpeg commands can definitely get thorny.

Since your README.md starts with an example of ffmpeg and gripes "This is not a normal sentence", it would be nice to simultaneously show the precise analogue as "ffhuman".

That way the reader immediately understands what your tool has to offer. Otherwise it can feel like we're learning two things instead of just one.

I'd say because the gaming business is entertainment; cyclical, discretionary and mostly dispensable. So, the makers don't really care to invest in optimizations like this. They spend a lot of money on marketing and artwork instead.

The Earth is a bitch, we've finished our news Homo Sapiens have outgrown their use

Oh, you Pretty Things (Oh, you Pretty Things) Don't you know you're driving your Mamas and Papas insane? Let me make it plain Gotta make way for the Homo Superior

   --- David Bowie circa 1971

You've posed this as a loaded question. Before asking it might be wise to first establish that there is an AI bubble (and that it pops).

I liked this one. Even though it's older I did see in 2025 myself. Does that count?

https://www.youtube.com/watch?v=yTicYJDT1zE