input_sh

That's kind of the fault of grant givers.

Nobody wants to give you a huge grant to continue doing what you're already doing, you only get grants for doing something completely new and grandiose. If you're lucky, you may get like 20% of the grant to cover your other expenses that don't come with a caveat of having to spend it on something completely new.

Silly old me, how could've I forgotten about such drastic improvements between say Sonnet 3.7 and Sonnet 4.6. It's 500x better now!

Thank you for teaching me, AI understander. You're definitely not detached from reality one bit. It's me, obviously.

The kid can learn and become better over time, while "AI" can only be retrained using better training data.

I'm not against using AI by any means, but I know what to use it for: for stuff where I can only do a worse than half the population because I can't be bothered to learn it properly. I don't want to toot my own horn, but I'd say I'm definitely better at my niche than 50% of the people. There are plenty of other niches where I'm not.

And those mediocre engineers put their work online, as do top-tier developers. In fact, I would say that the scale is likely tilted towards mediocre engineers putting more stuff online than really good ones.

So statistically speaking, when the "AI" consumes all of that as its training data and returns the most likely answer when prompted, what percentage of developers will it be better than?

Even speaking from a pure statistical perspective, it is quite literally impossible for "AI" that outputs world's-most-average-answer to be better than "most engineers".

In fact, it's pretty easy to conclude what percentage of engineers it's better than: all it does is it consumes as much data as possible and returns the statistically most probable answer, therefore it's gonna be better than roughly 50% of engineers. Maybe you can claim that it's better than 60% of engineers because bottom-of-the-barrel engineers tend to not publish their works online for it to be used as training data, but for every one of those you have a bunch of non-engineers that don't do this for a living putting their shitty attempts at getting stuff done using code online, so I'm actually gonna correct myself immediately and say that it's about 40%.

The same goes for every other output: it's gonna make the world's most average article, the most average song in a genre and so on. You can nudge it to be slightly better than the average with great effort, but no, you absolutely cannot make it better than most.

Quite frankly, if AI can write better code than most of your engineers "hundreds of times", then your hiring team is doing something terribly wrong.

You can make it a nice round 2000 these days.

I never said Google isn't the problem, what I said is that going to an agency isn't gonna fix that problem any more than running a SaaS tool yourself will, because they're not Google and they have no insight into what Google made one website prioritised over the other. Because, as you've pointed out, Google is the problem.

> I have to make a full blown website just to ensure I don't get SEO-spammed into oblivion?

No, I said a crappy one on purpose. How good is it doesn't matter, the sooner the Google knows about the domain, the better. Might as well be a copy of your README file using one of the million SSGs GitHub supports that will turn that README file into a website. The only thing that matters is that the website exists and that Google knows about it before the other one.

That's why many people purchase the domain on day 1 before they even start building the thing and also why many have like a dozen domains in their account that is like a boulevard of broken dreams there to remind them once a year they haven't done anything with them.

Still cheaper than a SEO agency or in most cases even one month of ahrefs access.

> This is very generous of you!

No it's not, it's a sales pitch that intentionally ignores some of the things pointed out in the article. The author has invested time into proper SEO optimization, legit websites already link to it et cetera, it's all explained in the article.

From the perspective of a spammer: They need like 2 million MAU to earn below minimum wage. You're never getting those figures by doing something legit and actually useful to a tiny subset of people. You either need a vague site beyond any point of usefulness to anyone or you need a network of knockoff sites. The reason you can't compete with these shitty SEO spam version of your site is because they already have a network of "authoritative" (in Google's eyes) sites and all they have to do is to link from them to a new one to expand their shitty network.

From the perspective of SEO agencies: They can't guarantee results. They can tell you vague, easily-googleable best practices and give you an output of some SEO SaaS that's far too expensive for an individual to purchase. Ahrefs(.com) is the prime example of this, the cheapest paid version costs $129/month. Do you care about SEO that much? No, so you go to these agencies and give them money for them to give you the output of such a tool. But that SaaS also only contains vague and nebulous "things to fix" to follow "best practices" because they also cannot know what drives traffic to your competitor from the outside perspective.

My best suggestion would be to start a website from day one. Doesn't matter how good the website is at first, Google favours sites that exist for longer. If you're creating a website after the knock-off version already exist, you might as well give up immediately, it's gonna be near impossible to recover from that.

I want you to look at that summary you've posted again, specifically the TRIPS part of it. Trade-Related Aspects of Intellectual Property Rights only came into force in 1995. Is that not the evidence that the rest of the world was strongarmed into adopting similar laws? The EU, Japan and a few others definitely supported the US in this initiative, but it was the US that heavily lobbied for it. Before TRIPS came to be, copyright effectively didn't exist in most of the world (yes, Western Europe excluded).

You're not contradicting anything I'm saying, you're just saying the same thing with more words because you don't understand where my argument comes from. What I don't understand is why you would even want to take credit for such a broken system to begin with.

> You are talking 0.05% if you are lucky.

That's... completely fine? One of my biggest pet peeves on this forum is someone like you mentioning half a million devices sold annually and somehow simultaneously calling that a failure.

You don't have to take over the whole market to be a successful company, many companies would be perfectly happy with selling half a million devices every year (AKA 0,05% of over a billion smartphones).

No I do not think it's an original concept of the US, more that it was the US that conditioned many other countries to adopt similar laws as a condition for trade deals / investments.

As a concept it existed in one way or the other pretty much ever since the printing press.

Life + 70 has always been an oversimplification, we still haven't even reached 70 years since the introduction of these rules (1973 in the US, in other countries depends on when the US strogarmed them into adopting similar rules).

There's all sorts of quirks for anything published before that rule got standardised more-or-less worldwide, but in general 1930-1945 is still like a legal grey area that can be challenged in court and you should be good to go for anything published before that. And don't get me even started on posthumous publications, that's a whole different can of worms where a family member might claim some contribution (like for example Anne Frank's father), pushing the copyright further to the life of the author + life of that family member + 70.

Look at the video, it explicitly starts with dealing with tasks in Markdown files. Unintuitively if I might say so, but it does.

I agree that now that this is a possibility, some sort of a wrapper would be great to see.

That 60 day limit was ignored so frequently in the past it might as well not exist.

Pretty much every attempt at stopping the president (from Clinton onwards) ends the same way: house votes on it, senate might agree with the slimmest of majority, it reaches the president's desk, president vetoes it, it goes back to the senate where it needs 2/3 majority to overthrow the veto, and it never gets that 2/3 majority.

The vast majority of all of those casualties throughout the years is the poorest rural population not being able to afford wood for heating, so they try to get some from the forests in their areas that nobody lays a claim on. You can explain pretty much every uptick by looking at how well the country's economy is doing.

Another common reason for some of the upticks are natural disasters (most frequently floods or landslides) hitting known minefields, which then makes the position more difficult to assess.

There's also a very particular type of mine that doesn't explode on the spot, but shoots ~1m into the air before exploding and sending shrapnel in a large radius which is by far the most deadly (https://en.wikipedia.org/wiki/PROM-1). The reason I'm mentioning it is that there was one particular incident with 5 casualties at a scrapyard that year. That same mine is also responsible for the only casualty last year, when a 19-yo goat herder tried to get his goats back on the road.

What you can do is look at the casualty rate per year and see how it went from hundreds in the 90s post-war period, to dozens throughout 2000s, to 0-1 in recent years and calculate your likelihood. Here's that chart for Bosnia: https://en.wikipedia.org/wiki/Land_mines_in_Bosnia_and_Herze...

I lived along former front lines my entire life, I spotted some unexploded ones with my own eyes a long time ago, I'm not dead yet. That said, I am more equipped to handle such situations than you are, as I was taught how to do that since my first days in school. I firmly believe that outsiders are frequently overreacting these days, which is completely understandable, but that doesn't mean it is rational. Your odds of winning a lottery are infinitely higher than dying to a landmine in the few days you spend here. You can increase them by doing something stupid like avoiding those markings, but besides that you'll be fine.

Bullet hole areas != land mine areas.

Think of it this way: bullet holes are where the fighting took place, while front lines have fluctuated. You don't want to mine an area that your soldiers might want to advance through. Land mines are placed when front lines have stabilised (like they are right now in Ukraine) to prevent the other side from advancing through. You only do that once your side has no intention of advancing further.

As such, land mines were usually properly documented and clearly marked as such after the war with giant skulls and red tapes, usually with some combo of words "PAZI MINE" ("beware, mines"). So while there are still rural areas that are littered with bullet holes, that does not mean those same areas were full of mines. It's also highly unlikely for a mine to be on any road, especially if it looks fairly well-maintained. You can take a road going through the minefield just fine, but you should never be one of those urban explorers that intentionally strays off of the road to look at the ruins on the side of that road.

It means there are no known areas that are still littered with landmines, but yes, that's not a guarantee there aren't any.

Not Croatian but Bosnian, 2030 is our target for this milestone and we have to keep de-mining ~70 square kilometres every year to be able to hit that milestone.

Only in America, in the rest of the world Google calls it "Gulf of Mexico (Gulf of America)".

I think that tells you more about the uselessness of SOTA benchmarks.

I can, easily. Speaking for my country and assuming you are not on a fixed-term contract, you can only get fired for one of the three reasons:

1. Company's in financial trouble and forced to downsize.

2. The position becomes obsolete and there's no option to transition you into some other role. In this case, the company can't hire anyone with a similar-enough skillset to yours for at least a year (or maybe even longer, I'm not sure).

3. Gross incompetence, in which case you need to be given an opportunity to course correct via a few documented warnings before being fired. Every warning requires your signature so that the company can't just make them up and backdate them.

That said, you don't become a permanent employee on day one, a company can issue up to three fixed-term contract before being forced to give you a permanent one. If you're on a fixed-term contract, they can just not extend it without having to satisfy any of the criteria above. But after a maximum of 3 years at the same company (as the maximum length for a fixed-term contract is one year), the criteria for firing you increases drastically.

So, the only way this could happen in my country is if the company stops renewing fixed-term contracts for recent employees, but then it wouldn't all be at the same time and you'd get the hint before the time comes to plan accordingly.

IF alias NOT ON allowList MOVE TO specialLittleFolder

By far the worst one is always going to be something generic like contact@, but my email provider is very good at filtering out those appropriately. :)

A simple regex filter will get rid of that. Now, if you use your own domain and have it configured as a catch-all, then you could do github@domain.tld.

May not necessarily be from commit messages, there's at least one way simpler way: simply adding .gpg to the end of any user URL will return that user's public GPG key.

Astro has "server islands" which rely on a backend server running somewhere. If 90% of the page is static but you need some interactivity for the remaining 10%, then Astro is a good fit, as that's what makes it different than other purely static site generators. Unlike Next.js, it's also not tied to React but framework-agnostic.

Anyways, that's why it's a good fit for Cloudflare: that backend needs to be run somewhere and Astro is big enough to have some sort of a userbase behind them that Cloudflare can advertise its service to. Think of it more as a targeted ad than a real acquisition because they're super interested in the technology behind it. If that were the case, they could've just forked it instead of acquiring it.

From Astro's perspective, they're (presumably) getting more money than they ever did working on a completely open source tool with zero paywalls, so it's a win-win for both sides that Cloudflare couldn't get from their vibe-coded project nobody's using at the moment.

> The AI boosting from the likes of you

Again, look at my comment history. I'm not discussing AI-as-a-whole because as you've pointed out it's not the topic of this discussion. I'm discussing how trivial it is to turn off as opposed to literally anywhere else, and that's not even discussing the provider choice you don't get anywhere else.

There's a whole section in macOS/iOS settings titled "Apple Intelligence and Siri" with ChatGPT being the only option, and you're seemingly happy with that compromise. Yet here you are complaining about an integration that's even easier to turn off and allows you to pick between 5 providers. There is literally no way of triggering it that doesn't immediately show you the "turn it off" button as it is right now (as in before this update reaches me).

I also invite you to go to firefox.com right now and find me a single mention of AI, since you for some reason are imagining that it is being advertised as an "AI browser".

"You people"? Take a look at my comment history and see my takes on AI please, but this is like the least harmful way of integrating it and yet "you people" are the loudest about it.

Can you do the same on Windows? Is it tucked away in settings on macOS? Can you disable it on Google? Can you disable it anywhere else? Why are you the most vocal about the integration that is literally the easiest to turn off? You need two clicks to do it right now, you're gonna need at least three once this kill switch is in settings.

Then you right-click on the AI button and click on "remove", but that's a whole different discussion than what you asked in the previous comment.

It's also why I really don't understand the need for a kill switch to begin with (other than pleasing annoying users), you don't need to wait for it. You can already get rid of the chatbot integration, there's a remove button already. It's also kind of annoyingly easy to misclick it, so they're just gonna remove it from those places and put it away in settings and those same annoying users will consider that a win.

IMHO no. Every chatbot has so much wasted space, it really doesn't need to be full-width. Also, what's easier?

Option 1: Being on a tab, copying the URL of the tab, switching to the chatbot tab, pasting the URL and writing some instructions about what to do with that tab.

Option 2: Clicking on the "summarise page" button (whether from the sidebar or from right-click context menu), and having the browser pre-fill the prompt with the URL + the reader view version of the content on that page.