dmayle

I think your math is a little off (or maybe mine is).

I'll take a short cut and imagine that you have an 8x8 square with no margins (68% of a borderless 8.5x11), then you have a grid of 600x600 bitmaps, which is 3.6e5. if each pixel is only black or white, than you have 1.8e19 possible bitmaps (64-bit), divide the two and you have 5e13, or about 50 trillion pages. Fix the equation, and you get a grid of 5.2e5, for 30 trillion pages instead of 50.

However, bring that up to 24-bit color or more (even 8-level greyscale is e154), and the exponentiality of the problem goes back to as described by the OP

I answered my own question! It's called Rogue Moon (1960) by Algis Budry, and was in the anthology The Science Fiction Hall of Fame, Volume Two

On the vein of similar books, in the late 80s, early 90's a read a science fiction anthology that had more or less the same exact story as Mickey7.

Humans discover a place (I think on Mars), built by Aliens, but it's a deathtrap. So they send someone in to navigate the deathtrap using a clone, and a sort of remote control (something like Avatar).

Each time the clone dies, the person piloting it survives, but has gained the memory of what went wrong, and can try again (kind of like Edge of Tomorrow).

The point of the story is the very end, when the pilot makes it fully through the space.

Has anoyone every heard of this storay and know the name/author? (Bonus points of you know the anthology as well)

There's no incompetence involved.

If you read the story, none of the more expensive, "legacy publishing houses", like Elsevier, had their contracts cancelled.

All of the publication access to university presses and open access publishers were cancelled.

This is another power/money grab

It's the scope. Executive orders only apply to the execution/actions of the federal government. Laws apply to the general public. The creation of laws is a legislative power, and reserved to the legislative branch of the government.

An official language isn't a mandate that enforces the use of English (take as an example India, which is probably the most multi-lingual country there is, but has Hindi as the "official language of the government". It's not the "national language", though in their case the splitting of hairs is probably directly tied to the number of languages spoken).

I know an American who spent several years in Paris, working in English, and never learned more French than was necessary to order her coffee. And this is despite French being the national language of France. (Interestingly enough, France is very multi-lingual as well, with many regions having their own history and language like Provençal, or Niçois, Breton, etc.).

It's more about a recognition of the status quo (as another commenter pointed out, most people acquiring US citizenship are required to demonstrate the most basic level of proficiency with the english language), and ensuring a baseline of support, no matter where you are in the country. Multi-lingualism is a plus, and in no way hindered by having a national language

There's an effect to be sure. I am, however, talking about legal status. As one of the other commenters pointed out, the actual text of the executive action is about rolling back a previous executive action designed to prevent Title VI discrimination. So, if I were affected as mentioned in the parent post to this, I would be reaching out to the ACLU to find a lawyer to help with challenging this.

As to the stated wording though, "making English the country's official language" (which mirror's wording used in the most recent presidential address), the executive order has no power to do so, because it's not a law.

> I guess it depends on what that means.

A quick Google suggests that about 22% of Americans are multi-lingual, and roughly 8% of Americans do not have English as their dominant language (though I suspect that entirely non-English speaking Americans are at least one order of magnitude less).

I think only about 2% of countries in the world don't have an official language (America among them), and yet about 50% of the world is multi-lingual, so having an official language doesn't seem like an obstacle for other governments when it comes to support.

I suspect that the lack of of national language has more to do with the power struggle between the federal government and state governments then any issue, which is why I find this latest executive order quite baffling (for a party that is in the process of dismantling federal government, this is very much a federal power grab).

With all of that being said, I see it both as a recognition of the status quo, and a commitment to what is one of the greatest strengths of the United States (a single-language, single-currency market). Our largest economic rival (China) also has an enormous single-language, single-currency market, but that strength is largely focused inwards, since the use of a non-romanised language makes it very difficult for the non-Chinese reading population to adop sub-parts of the Chinese language.

Is it just me, or is what OpenAI is really lacking is a billing API/platform?

As an engineer, I have to manage the cost/service ratio manually, making sure I charge enough to handle my traffic, while enforcing/managing/policing the usage.

Additionally, there are customers who already pay for OpenAI, so the value add for them is less, since they are paying twice for the underlying capabilities.

If OpenAPI had a billing API/platform ala AppStore/PlayStore, I have multiple price points matched to OpenAI usage limits (and maybe configurable profit margins).

For customers that don't have an existing relationship with me, OpenAI could support a Netflix/YouTube-style profit-sharing system, where OpenAI customers can try out and use products integrated with the billing platform/API, and my products would receive payment in accordance with customer usage...

One of the starting points of this article is that the current president has signed an executive order "making English the country's official language".

I think it's important to remind people what executive orders can and can't do. An executive order is an instruction sent to the government itself. It instructs government workers how to perform their job. It is not directed at the American public (though it can and does have an effect on the American public by way of government policy).

As such, this current executive order effectively does nothing. We've attempted to pass laws that make English the national language, but have consistently failed to do so.

And personally, I'm for having English be the national language of America (as a bilingual American myself), but this executive order does not make that so.

So the story originally started with the cacheless 266 Mhz Celeron. CPUs were delivered as AICs (add-in-cards) at the time, with separate cache chips, so to deliver a budget processor, they shipped the same silicon, but without the cache chips added. Removing the cache drastically tanked the performance, especially on integer work loads (typically productivity software), but didn't really affect floating point workloads. However, it had the side benefit of removing the part of the AIC that was most sensitive to over-clocking (the cache). It used a 66Mhz clock with a fixed 4x multiplier, and upping the clock to 100Mhz got the Celeron running at 400Mhz, which had performance roughly equivalent to a 266 Mhz Pentium II with cache for integer workloads, but for games, it was almost as fast as the fastest Pentium II of the time (which topped out at 450Mhz).

In order to stop the overclocking, Intel decided to add some cache back to the CPU, but to save money, rather than using cache chips, they stuck a relatively tiny amount of cache directly on the CPU die, and released the now infamous Celeron 300A

Because the cache was on-die, it could overclock just as well as the previous celeron, but this time the 300A was faster than the equivalent Pentium because the on-die cache ran at twice the clock speed of the external caches

It's called nominative determinism: https://en.wikipedia.org/wiki/Nominative_determinism

This is why H1-B visas should have a minimum salary requirement equal to 20% over whichever is greater, median salary for the role in the industry, or median salary for the role in the company (and whichever is greater, US-wide, or local pay scale).

This way, a company is always incentivized to find local talent, but when they are actually unable to, they have a path to find the expertise they need. The U.S. could relax restrictions on H1-B, lowering red tape, and removing a lot of churn that comes with the H1-B program

You actually have it backwards. Your claim is that legacy admissions bias in favor of the predominant race might be true for a school that had race-blind admittance criteria. In the opposite case, however, legacy admissions bias against people of the predominant race (for the general student).

Since legacy admissions come first, schools which practice affirmative action have a heavy bias against the predominant race (because those slots are all filled by legacy candidate). Which means that if you're of the predominant race, you have next to no chance to be accepted by these universities... (I mean, everyone has next to no chance, but for people of the predominant race, they are discriminated against severely).

In general, though, college admissions are pretty terrible... Having spoken with someone who worked in admissions at one of these universities, if you have a bright kid, you're better off moving to the middle of nowhere to make sure they're the valedictorian, rather than trying to send your kid to a great high school where why might only be salutatorian. Why? For smaller schools they rarely take more than one student from that school in any given year, so when the valedictorian who filled out applications to 10 top schools gets in to all 10? The salutatorian doesn't...

It's not a bribe, because the government claimed that Google owed this amount of money, and Google paid it. If this was a check to prevent Government action, you might consider it a bribe. If it had been paid to an individual, you might consider it a bribe. In this case, the law says that there can only be a trial by jury if there is financial damage. With Google conceding the amount claimed and paying it, there is no financial damage, and thus there can not be a trial by jury. The money is not being spent to purchase a trial without jury, it's a fine being paid to change the rules of the encounter.

Except Isreal isn't a signatory nation, and all of the land in question is part of Israel. That's not a judgement either way, there have been some attempts to change that, but the ICC doesn't actually have any jurisdiction here.

From personal experience as far back as the 80's (and from my understanding going back before that as well), OS has never meant kernel.

An Operating System is the collection of software that allows you to operate a computer, so that means kernel, program loader, simple text editor, simple disk management, etc.

As computer users became more savvy, and hardware became more powerful, more and more functionality was included in the OS (graphical interfaces, utility apps, etc.).

I don't think many people would have trouble calling Android an operating system, and that's just the Linux kernel with utility apps, loader, and app libraries, yet very different from something like Redhat.

I don't think it's a stretch in the least to call Horizon an OS.

I started reading this, and the article is very approachable, but it’s flawed from almost the very beginning.

He uses the angle addition formulas to derive the Pythagorean theorem, but this derivation only works because he predefined sine and cosine as odd and even functions, which hasn’t been proven

That's actually not true. It can do nothing about M of N cryptography. (That's when a key is broken up such that there are N parts, and at least M (less than N) are required to decrypt. It doesn't matter how many rubber hoses you have, one person can fully divulge or give access to their key and it's still safe.

I've just started down that route. I've got the nitro key hsm2 in the mail, have heard the advice on using two levels (first root in the Key, and intermediary on the Device for easier revoking). I mainly want to issue client certificates so that I can expose internal sites on the public Internet via proxy without having to require a VPN for all of my users, though I'm also interested in certificate based SSH

Are there are books on the topic, or at least good tutorials? I know the underlying concepts (have even read most of the specs), and the documentation has gotten better since I last looked at it, but I'd love to have a more comprehensive resource.

I wish that Framework would respond officially as to whether or not they will support ECC on the AMD boards. I make it a point to buy AMD processors and AMD motherboards that support ECC. I've seen the bitflips (not often), I know they happend, and I appreciate the added security, especially given how little of a cost difference (about 12% in terms of BOM for just the ram, or 1-2% of the whole system cost)

This is the most ridiculous take on this that I've ever seen. Next you're going to say that newspaper stands need to pay a charge to the newspaper each time someone walks by their stall (or buys gum, for example). They have seen the headline, and then moved on.

Very often a headline is all someone needs or wants. That has value. Without anyone buying that newspaper, the newspaper stand has got value from the headline, and contributed nothing to the newspaper publisher in return.

The reality is that headlines are advertisements for articles. That's why there are headline writers in the first place. Make a better advertisement, get more sales.

In the case of Google, publishing links with headlines means publishing free ads for that website. The website most certainly benefits from that relationship, if they didn't, they would just use robots.txt to block Google indexing their website, which someone has always been free to do.

The real problem is that newspapers would just like to take a percentage of Google revenue, because they're a big company.

The paper has a fundamental flaw, which can be seen if trying to reverse the reasoning to look at the past, rather than trying to predict the future.

If you used today's level of computation with 2012 era models, then you wouldn't have the error rates of today's models, you would have much higher, much worse errors.

The single biggest bottleneck in deep learning is not computation, it's the ingenuity used to devise new structures and new optimizations that allow for scaling.

For a given structure, with given techniques, you can throw more computation at a problem to decrease error rate, but the gains scale poorly with cost. Cost improvements only come with new techniques and structures.

When I originally brought it up, no one was interested in the idea. It looks like this was introduced about six months after I left (which is about 18 months after I first brought it up), so it looks like they changed their minds.

I am not an Apple fanboy (I run Nix on x86 mostly, some Arm), but I actually think they've managed to get it completely right on this. There are a few key pieces that I see as signs that this will eventually hit it out of the park. With that being said, I do believe this is a v1, but remember that the first iPhone was very much a v1 as well.. (I stayed out until the iPhone 3GS myself, and prefer Android phones nowadays). The signs:

1. The battery seems replaceable. I had an original Google Glass... after 2-3 hours of use you had to take it off your head and let it charge, meaning that you would never get used to always having it on. If people have to choose between having a useless hunk of heavy metal on their head, or not being able to use it while it recharges for a few hours, you have failed. 2. Making it AR-first and having that weird external display. Real life will intrude. Can you answer the door, turn off an appliance, do anything else, without having to pull something off of your head and unstrap some controllers from your hands? With the Apple model of handling things, you can pop out to real life without having to get in or out of using it. If you have to take it off, you're just a little bit less likely to put it back on. 3. Tapping your own fingers to click. This was something I brought up to FB/Meta when I was working there, but they decided not to use it. The only way you can give a tactile/haptic response to virtual items, is if there's also some sort of physical item. Tapping your own finger is the single easiest thing that makes sense. I hope that Apple adds future support for augmented 'props'. Imagine an empty artists palette (typically white plastic), that has virtual colors in swatches, or in a color wheel. You can tap on the physical palette to choose your color, to virtually mix the paints, etc.

"If your badge has been hit by ransomware, see how to enter Safe mode."

What a world we live in...

There might be another way to look at it that could be really useful in this scenario. Rather than saying that you don't want it to happen, instead, ask yourself what changes would make it so that you would be okay with the change? For example, if the CUXO was willing to sign over half of his stake (10%) to you, would you be okay with him taking the CTO title? You could argue that giving him the CTO title will give him a future leg-up in career and earnings that you should be compensated as well. If the CUXO is not okay with that, than you have a good argument for keeping things the same.

Fun talk... In my mind, I've always seen Don't Repeat Yourself, and thought the opposite was Write Every Time.

Hey! I pre-ordered an AMD version as soon as the site would handle it, but I wasn't able to choose between the mattte and glossy display. Is there any way to get the matte display with the new AMD option?