LeifCarrotson

You can't view a random imgur screenshot in the UK?

What image hosts work for you? Imgbb? Postimages?

In my experience, a spec is often public (or available for purchase by the public from a standards committee for a few hundred dollars, close enough in business terms), but the source code for your competitor's implementations is not available for any price. The first barrier to entry to become a competitor in many industries is to implement an 835 page (note that those are rookie numbers) spec, building software that is compatible with existing data formats or legal frameworks or network protocols or hardware or whatever.

This is another example of the many invisible silos that developers end up in:

1. Some are developing enterprise software for use by their own employer, the spec only exists because it's useful to help the dev team manage the project. In this case, it's totally reasonable to expect the spec to co-evolve with the code. If the release gets pushed out to Q3 or the target has to be adjusted due to unforseen implementation difficulties, so be it.

2. Some are developing contract software for their customers. The spec exists as a communication tool and requirements that allows the developers to know what to build and to inform the customer that what they've built has reached a certain phase of the billing process (including completion). You're on a clock, and if you're late or you don't do what you said you'd do then your customer isn't going to pay you.

3. Some are developing software for abstract customers - either B2B or B2C. Often, the specs here are regulatory or industry-standard, implementing the spec is the function of the business. You're releasing updates and new products to make year-end sales targets or snipe announcements by your competition, and advertising compatibility with various standards is a crucial bullet point in the sales presentation.

Bug reports, feature requests, and especially deadlines have different meanings in each context. The way AI development agents can be used in each silo also varies enormously!

If you're willing to put a little time into video editing, a 360 cam is great. The insta360 tools can make that a little easier if you want something simple.

If you just want to store a snapshot of the moment as it was captured, a regular camera that you pointed in the right direction is better.

Low power electronics are all about low duty cycles: if it can be worn but asleep and drawing next to no power for 15 hours and 59 minutes, and recording audio for 60 seconds, they only need enough battery to run the processor for 60 seconds.

No one is using the thing like a dedicated microphone to record their entire day's audio, it's a note-taking tool.

On the other hand, reducing competition and consolidating the market within the combined organization is good for the misshapen franken-org that comes out at the end...

The question is whether or not they can weaponize the regulatory system to prevent smaller competitors from taking advantage of their confusion.

I think that any workflow that requires the user to stare at the tokens being generated live is using it wrong. Delegate, don't stare!

https://mikeveerman.github.io/tokenspeed/?rate=10&mode=text

You think of an idea that you want to have the LLM process, queue it up, and go back to what you were doing. Once you've finished reading the next article on HN about a 5 tps Xeon, your task will be complete. It's kind of like using a 3D printer: It doesn't matter if a print takes 10 hours, because when you come back in the morning it will be done.

Yes, with top-tier GPU farms you can hit hundreds of tokens per second. But if the old Xeon in the closet can get useful work done at 5 tokens per second, there are lots of people and lots of use cases where a free, unlimited 5 TPS stream is worth more than paying a dollars per day to get access to a 500 TPS source.

Humans easily anthropomorphize things that are not humans, ascribing human attributes like motive and comprehension and emotion to objects and processes that are not people who can have those attributes.

Claude is not a human.

It is overwhelmingly easier to anthropomorphize Claude or Siri or an LLM that communicates with you more eloquently than your boss than it is to anthropomorphize a cranky, tired starter motor. It's often easier to do than it is not to do, and sometimes, it's a useful abstraction. But it's not precise or correct, and can result in errors.

It could also just be that they're getting confused when using tools configured without a username dedicated to the tool. It's easy to end up with a comment or commit message that says "I prefer X over Y" posted on Alxndr's account and have coworkers confused whether that's the LLM or the human making that statement.

Tractors are immortal (my Dad's 80-year-old Ford 9N can attest to that), John Deere is immoral.

I'm not left-handed, but I often scroll my phone with my left thumb. My right hand is on my computer mouse, or holding a pen, or employed to make precise touches on the phone screen with my right index finger, or briefly comes over to join with my left thumb for typing...

Scrolling doesn't require much precision, and I naturally hold my phone in my left hand.

Your terminal doesn't typically use Consolas. You can do:

    $ wget  https://archive.org/download/PowerPointViewer_201801/PowerPointViewer.exe
    $ cabextract PowerPointViewer.exe
    $ cabextract ppviewer.cab
    $ open CONSOLA*.TTF

What they're suggesting is that a lot of their users are logging into an Akamai Linux box from a Windows machine, and therefore aren't "real Linux geeks".

It's 0.0015, that's 1.5 to 2.5 thousandths, or 15-25 "tenths" as they're called.

That's not a particularly tiny gap in the machinist world, it's large so that you can pump viscous oil in it and deal with a wide variety of temperature changes.

25 thousandths would be sloppy, a nominal clearance hole for a 1/4x20 bolt is about that much.

Auto start/stop isn't off for enough time for the oil to drain from the galleries and especially not out of the bearing journals.

It's the first few seconds after an engine has been off for hours (or worse, for potentially years) that are the problem.

What they meant when they said "planets" was the 8 (previously 9, previously to that 8, previously 7...) known and named planets in our own solar system. A hypothetical "Journal of Planets" that was actually about solar system astronomy wouldn't necessarily have known what to do with a new paper about 51 Pegasi b published 30 years ago. They're thinking "when we said planets, we actually always meant solar system planets, it just never came up until now".

The reviewer of this paper is saying that by biology they always meant naturally evolved cellular biology, not synthetic biology - there's just never been an example of the latter before.

I think the take is wrong, the receiving journals should be excited to expand their scope rather than frustratedly redefine their scope more narrowly, but definitions and categorization are hard.

I work in industrial manufacturing and automation, several of my customers (those running steel foundries, aluminum die casting, plastic recycling and extrusion, and other power-intensive processes) represent a sizeable fraction of the utility usage in the small towns in which they're located.

They often have an individual contract with the utility and participate in load regulation: when you need liquefy a few tons of steel, those heaters have a lot of thermal inertia. If A/C loads are high they'll turn the power down, if wind output is high, they'll turn it up, and so on.

Do data centers participate in the same sort of dynamic pricing and power adjustment? I understand that they're spinning up and powering down instances on demand, and that those demands are somewhat outside of their control, but are they able (and willing, and desirous of reducing their electric bills) to dynamically adjust compute in response to utility rates?

I've also been running Qwen 3.6 35B A3b on my Windows laptop (64 GB RAM, a 4GB GPU) and it's at least tolerable. It's not fast - a few tokens per second, slower than reading speed - but I can give it a task and come back later. That was a $600 laptop off eBay a few years ago, not a $6,000 machine.

Are these unified memory Macs and giant 24GB desktop GPUs achieving dozens or hundreds of tokens per second commensurate with their 10x-20x cost?

You've described a load cell, which is commonly used in laboratory and industrial environments to resolutions exceeding 0.05% of full scale. I've used these in the past with great success:

https://www.interfaceforce.com/products/load-cells/low-profi...

Click the "Specifications" tab to see the various error sources. Interface's technical library (linked at the top right of their shop) is a great training tool too, if you want to learn more. To keep your NIST certification - and to maintain accuracy after a high-stress...incident... you need to do occasional recalibration. That would be a problem for a consumer product, but with good mounting and overload protection they really don't change that much over time.

Then you've got the electronics, and the amplifier required to get the rated precision out of that load cell costs more than the load cell itself.

Achieving those performance targets costs money, and people buy the cheapest scale that says "Accurate High Precision Digital Bathroom Scale Glass" on the Amazon listing.

Yes, that's "bare metal" as opposed to programming in MicroPython, Javascript with Espruino, Lua/NodeMCU, Rust, or adding your C application on top of an off-the-shelf RTOS.

C support is basically universal these days, very few chips require you to program in only assembly anymore.

"They" being the waiter or waitress, of course. Good ones can navigate complex bill splitting arrangements and even better can manage awkward interactions like one person quietly paying for another's dessert or covering an appetizer for the table, know the menu not only by memory but also can recommend dishes that the guest may prefer, and generally make the dining and ordering and paying experience better.

Bad restaurants think they can replace those skills with a QR code on the table optimized for the lowest common denominator.

I don't think they're necessarily overestimating what they've dedicated to logistics, they're overestimating what they've dedicated to anti-scalping. If they cared to detect it, they could invest even just a token effort and make great strides. But why would they care?

From the paper:

The main sources of VMS were determined to be antiperspirants ... skin lotions ... wipes ... and hair conditioner. Several siloxanes-free options are available for [these products]. These products are now being assessed for crew member use in future increments.

From the blog:

At present the agency is testing a new filtration system to put in front of the heat exchangers, to try to protect them, and continuing to try to cut down on siloxanes at the source level. There are probably people at NASA now whose entire career has been built on siloxane control.

Why wasn't the result to simply ban siloxane-containing cosmetics and wipes? The cosmetics are up to the individual astronaut, which is a little crazy, but the wipes are provided by NASA, and they're still using siloxane-bearing wipes, which shortens the life of their water systems and costs crazy amounts of money.

Those genetic requirements come into play at elite levels, but you need to start young, when those differences are less obvious.

You need to look at what sports an eight-year-old is playing in the backyard, what sports his Dad is excited about on the TV.

An agile, fast, coordinated kid who's coachable and wants to train hard but is going to grow up to be 5' 8" is not going to make the NFL or the NBA, but if they've got the athleticism to play in the World Cup... well, in the US that kid will be the point guard on the local high school basketball team and also play safety and wide receiver on the football team.

In India, they'd be a cricket star.

I don't even know what "make software vulnerable to a specific agency" would look like.

Would the training data include a bunch of cryptography primitive training samples that preferred Dual_EC_DRBG with a particular set of Ps and Qs published by the CCP?

Are you 20? I got 35 unc which is a few points less than my IRL age.

I'm curious at what (if any) age the mean score skews to greater than age in years. I expect there are multiple brackets, the very young won't check off many at all but by some of the entries are sufficiently niche that when an octogenarian takes the test they won't hit more than 80% of the options.

Reminds me of the opening to "Ministry for the Future" by Kim Stanley Robinson.

In that book, a wet bulb event (high humidity and high temp) in India pushes infrastructure past the breaking point, the grid goes down, AC systems still running on generators are overloaded and overcrowded and fail, the water temp goes over body temp, and millions die.

The positive cultural/societal reaction to the disaster strained my suspension of disbelief pretty hard, as is typical of KSR novels in my experience, but the idea of a heat wave causing a massive catastrophe (and the poignant description of attempting to live through it) stuck with me.

Content creating is not maintenance. And hosting of a static site is dirt cheap, caching works flawlessly.

It could've been a two-million-dollar website if he'd tried to roll his own CMS and Javascript framework, for zero benefit over the one-million-dollar website he actually built.

Many G-code dialects contain subprogram calls, loops, and conditionals that seem like they'd permit a physical quine. The standard RepRap firmware breaks the rules by taking input:

https://reprap.org/wiki/G-code#M98:_Call_Macro.2FSubprogram

but Fanuc and other dialects allow M97 to call a subroutine chained to the end of a program.

They come from lower parts of the patient's scalp. The typical "receding hairline" pattern is caused because the follicles on the forehead and top of the head become sensitive to with age to DHT, the but the hair follicles on the lower and on the back of the head don't have the same sensitivity. There's usually more than enough of these resistant follicles to maintain sufficient hair density.

The surgery just moves the follicles around your own scalp. Body hair transplant can be done but is relatively uncommon, donor hair from other people (or animals) requiring a lifetime on immune system suppressing drugs as with an organ transplant is virtually unknown.

No, it's not pubic hair and you don't need to have a hairy back or chest, and no, there are not millions of low-status Turkish men walking around with scarred heads because they sold their scalp to a foreigner.

Not "just", it's (presumably) 8 dedicated pins that form an RMII interface. This is not the same 8 pins as you'll find in your 4-pair Ethernet cable, it's a separate protocol which can be connected to an Ethernet PHY transciever like a TI DP83867E [1], which is further connected to "magnetics" [2], a convenient package of 8 integrated transformers and chokes that provide the galvanic isolation feature of an Ethernet connection.

A few SoCs provide integrated PHY transceivers, but usually it's an external chip.

[1] https://www.ti.com/lit/ds/symlink/dp83867e.pdf

[2] https://yageogroup.com/content/datasheet/asset/file/DATASHEE...

Yeah, but even when it was a new quote, it was inaccurate: People have always wanted to follow their own desires, always wanted to be told what they want to hear, and always disliked listening to harsh truths. Even in 64CE.

Much like planting a tree, the ideal time to use your own domain for email was many years ago, but the next best time to do it is today.

Do it! minraws.com (if that means anything to you) is available, you could be firstname@minraws.com as well as your @gmail.com before the sun goes down. Personally, I'd set it up to feed into a new mailbox with Protonmail, but if you like you can just have it forward to the familiar gmail inbox you're used to.

You can start moving your accounts over one at a time. It doesn't have to be instant. Yeah, there are probably IRL business cards in drawers and people you haven't contacted in decades that will mean that you want to forward all emails that go to your gmail to a folder/label in your new email domain forever, but that's OK.

Just start.