That's because there are at least 5 different definitions of AI.
- At it's inception in 1955 it was "learning or any other feature of intelligence" simulated by a machine [1] (fun fact: both neural networks and computers using natural language were on the agenda back then)
- Following from that we have the "all machine learning is AI" which was the prevalent definition about a decade ago
- Then there's the academic definition that is roughly "computers acting in real or simulated environments" and includes such mundane and algorithmic things as path finding
- Then there's obviously AGI, or the closely related Hollywood/SciFi definition of AI
- Then there's just "things that the general public doesn't expect computers to be able to do". Back when chess computers used to be called AI this was probably the closest definition that fits. Clever sales people also used to love to call prediction via simple linear regression AI
Notably four out of five of them don't involve computers actually being intelligent. And just a couple years ago we still sold simple face detection as AI
It's the opposite. It is doing the driving but you really have to provide lane assist, otherwise you hit the tree, or start driving in the opposite direction.
Many people claim it's doing great because they have driven hundreds of kilometers, but don't particularly care whether they arrived at the exact place, and are happy with the approximate destination.
Is the siren song of "AI effect" so strong in your mind that you look at a system that writes short stories, solves advanced math problems and writes working code, and then immediately pronounce it "not intelligent"?
It doesn’t actually solve those math problems though, does it? It replies with a solution if it has seen one often enough in training data or something that looks like a solution but isn’t. At the end, the human still needs to proof it.
Same for short stories, it doesn’t actually write new stories, it rehashes stories it (probably illegally) ingested in training data.
LLMs are good at mimicking the content they were trained on, they don’t actually adopt or extend the intelligence required to create that content in the first place.
Oh, I remember those talks. People actually checking whether an LLM's response is something that was in the training data, something that was online that it replicated, or something new.
They weren't finding a lot of matches. That was odd.
That was in the days of GPT-2. That was when the first weak signs of "LLMs aren't just naively rephrasing the training data" emerged. That finding was controversial, at the time. GPT-2 couldn't even solve "17 + 29". ChatGPT didn't exist yet. Most didn't believe that it was possible to build something like it with LLM tech.
I wish I could say I was among the people who had the foresight, but I wasn't. Got a harsh wake-up call on that.
And yet, here we are, in year 20-fucking-25, where off-the-shelf commercially available AIs burn through math competitions and one shot coding tasks. And people still say "they just rehash the training data".
Because the alternative is: admitting that we found an algorithm that crams abstract thinking into arrays of matrix math. That it's no longer human exclusive. And that seems to be completely unpalatable to many.
You and I must be using very different versions of Claude. As an infra/systems guy (non-coder), the ability for me to develop some powerful tools simply by leveraging Claude has been nothing short of amazing. I started using Claude about 8 months ago and have since created about 20 tools ranging from simple USB detection scripts (for secure erasing SSDs) to complex tools like an Azure File Manager and a production-ready data migration tool (Azure to Snowflake). Yes, I know bash and some Python, but Claude has really helped me create tools that would have taken many weeks/months to build using the right technology stack. I am happy to pay for the Claude Max plan; it has returned huge dividends to my productivity.
And, maybe that is the difference. Non coders can use AI to help build MVPs and tooling they could otherwise not do (or take a long time to get done). On the other hand, professional coders see this as an intrusion to their domain, become very skeptical because it does not write code "their way" or introduces some bugs, and push back hard.
Yeah. You're not a coder, so you don't have the expertise to see the pitfalls and problems with the approach.
If you want to use concrete to anchor some poles in the ground, great. Build that gazebo. If it falls down, oh well.
If you want to use concrete to make a building that needs to be safe and maintained, it's critical that you use the right concrete mix, use rebar in the right way, and seal it properly.
Civil engineers aren't "threatened" by hobbyists building gazebos. Software engineers aren't "threatened" by AI. We're pointing out that the building's gonna fall over if you do it this way, which is what we're actually paid to do.
Sorry, carefully read the comments on this thread and you will quickly realize "real" coders are very much threatened by this technology - especially junior coders. They are frightened their job is at stake by a new tool and take a very anti-AI view to the entire domain - probably more-so for those who live in areas where the wages are not high to begin with. People who come from a different perspective truly see the value of what these tools can help you do. To say all AI output is slop or garbage is just wrong.
The flip of this is to understand and appreciate what the new tooling can help you do and adopt. Sure, junior coders will face significant headwinds, but I guarantee you there are opportunities waiting to get uncovered. Just give it a couple of years...
Every HN thread about AI eventually has someone claiming the code it produces is “trash” or “non-working.” There are plenty of top-tier programmers here who dismiss anyone who actually finds LLM-generated code useful, even when it gets the job done.
I’m tempted to propose a new law—like Poe’s or Godwin’s—that goes something like: “Any discussion about AI will eventually lead to someone insisting it can’t match human programmers.”
Seeing an AI casually spit out an 800 lines script that works first try is really fucking humbling to me, because I know I wouldn't be able to do that myself.
Sure, it's an area of AI advantage, and I still crush AI in complex codebases or embedded code. But AI is not strictly worse than me, clearly. The fact that it already has this area of advantage should give you a pause.
Humbling indeed. I am utterly amazed at Claude's breadth of knowledge and ability to understand the context of our conversations. Even if I misspell words, don't use the exact phrase, or call something a function instead of a thread, Claude understands what I want and helps make it happen. Not to mention the ability to read hundreds of lines of debug output and point out a tiny error that caused the bug.
We use them in our shop. It's quite straightforward if you're already familiar with Github Actions. The Forgejo runner is tiny and you can build it even on unsupported platforms (https://code.forgejo.org/forgejo/runner) e.g. we've setup our CI to also run on Macs (by https://www.oakhost.net) for App Store related builds. It's really quite a joy :)
Are you building MacOS apps? More specifically, are you doing code signing and notarization and stamping within CI? If so, is this written up somewhere? I really struggled with getting that working on GitLab. I did have it working, but was always searching for alternatives.
I think the parent was being sarcastic, because while you can trigger some function on the AirPods themselves, a large list of features are unavailable, not documented and purposefully concealed (as the main post describes in detail). AirPods don't even report their battery life to third party devices (as normal Bluetooth accessories can)
It's telling that we can appropriate millions of dollars to transport a decommissioned shuttle from a museum in Virginia to Texas, but NASA can't pitch in the cost of one tank of diesel to the people maintaining what this article claims to be a mission-critical tool?
> Isn't identifying real humans an unsolved problem?
I'm not sure this was ever a problem to begin with. The obsession with "confirm you are human" has created a lot of "bureaucracy" on technical level without actually protecting websites from unauthorised use. Why not actually bite the bullet and allow automations to interact with web resources instead of bothering humans to solve puzzles 10 times per day?
> Cloudflare defining "trusted"
They would love to monetise the opportunity, no doubt
> I'm not sure this was ever a problem to begin with. The obsession with "confirm you are human" has created a lot of "bureaucracy" on technical level without actually protecting websites from unauthorised use. Why not actually bite the bullet and allow automations to interact with web resources instead of bothering humans to solve puzzles 10 times per day?
I mostly just let the bots have my sites, but I also don't have anything popular enough that it costs me money to do so. If I was paying for extra compute or bandwidth to accommodate bots, I may have a stronger stance.
I do feel a burden with my private site that has a request an account form that has no captcha or bot blocking technology. Fake account requests are 100 to 1 real account, but this is my burden as a site owner, not my users' burden. Currently the fake account requests are easy enough to scan and I think I do a good job of picking out the humans, but I can't be sure and I fear this works because I run small software.
I send them on endless redirect loops with very slow responses. Cost me very little bandwidth and it effectively traps one bot process that then isn't available for useful work. Multiply by suitably large 'n' and they might even decide to start to play nice.
>"Why not actually bite the bullet and allow automations to interact with web resources instead of bothering humans to solve puzzles 10 times per day?"
This is a great idea if you've developed your 'full-stack', but if you're interfacing with others, it often doesn't work well. For example, if you use an external payment processor, and allow bots to constantly test stolen credit card data, you will eventually get booted from the service.
I think the comment means we have these “institutional” problems that we’re constantly protecting with tricks like captchas instead of actually addressing why a payment processor would have a problem with that or be unable to handle it in their own way.
The average normal user would go months to years between needing to update payment info, so why would that require them to solve puzzles 10 times a day?
That is also notably a completely unnecessary dumpster fire created by the credit card companies. Hey guys, how about an API that will request the credit card company to send a text/email to the cardholder asking them to confirm they want to make a payment to Your Company, and then let your company know in real time whether they said yes? Use that once when they first add the card and you're not going to be a very useful service for card testing.
What you need is for all card issuers to be required to implement it by the network. Otherwise you'll still have people showing up to test all the cards that don't support it and the payment processors would still kick you off for that.
I tried Firefox just a few days ago, but it didn't work out. I just missed too many things out of the box. My main browser is Vivaldi (so all the chromium goodies + privacy + made in EU). Safari comes in as a close second, I tend to use it on the go because it syncs well with my Mac and Apple throttles any other browser on iOS.
I'm not sure making a list that's been done time and time again here would make a difference. But if you're truly interested, here are my top 3:
- Firefox comes packed with all kinds of telemetry and analytics turned on.
- No workspaces/profiles. I know there are extensions that can enable various flavours of this functionality, but it's just too much overhead to experiment and test every-single-one.
- Widgets! I love Vivaldi's Dashboard for when one needs more than just a homepage.
> No workspaces/profiles. I know there are extensions that can enable various flavours of this functionality, but it's just too much overhead to experiment and test every-single-one.
Most of it is built-in and these days there is generally only one recommended extension: Multi-Account Container. Containers (tabs in the same window with different cookie jars/etc) are built-in, the UI for working with them is not, and MAC is the generally agreed best UI. It lets you create as many named Containers as you want and assign them a color to gently stripe your tabs. You can right-click the new tab button and get a list of Containers to open the next tab in. You can right-click an existing tab and reopen it in a new Container. (TIL from this article there's an option to have left click on the new tab button always open the Container menu. I don't think I'd use that, but it's nice to know.)
If you want to go older school, Profiles (browser windows with different extensions/cookie jars/everything) have been around since the beginning of Firefox (and it shows in how old and ugly some of the UI still is, hah). `about:profiles` is a profile manager when you are already inside a Firefox browser. The `-ProfileManager` command line switch is the ancient startup option to manage profiles before opening a browser window at all. If you like to use side-by-side profiles often you want `-noremote -ProfileManager` and/or `-noremote -P $profileName` shortcuts. (`-noremote` says not to send it to any currently open Firefox window.)
I think the parent comment acknowledges that it's indeed possible but all that are so many steps one shouldn't have to perform just to get to the feature.
Firefox wants to be customisable, but instead of features, its out-of-box experience is focused on sharing activity with Google and giving you sponsored bookmarks... I agree that the priorities there are not exactly user-focused.
It's 1 (official from Mozilla) extension to install (easily, from a simple marketplace) or 1 url and/or 2 command line flags to learn just long enough to bookmark them/add them to a shortcut file somewhere.
I don't feel like that is "so many steps". Two choices, one easy thing to do whichever path you chose.
The one thing is an extension still because Mozilla still doesn't think the UX is right and doesn't want to include it out of the box until they think the UX is 100% solid, especially for general users, not just power users. The other thing has active improvements to its out-of-the-box UX already in beta testing/slow rollout.
The only activity Firefox shares with Google is if you leave Google the default search, which is a real easy to change at the top of the Settings page.
Sponsored bookmarks are dumb but easy to ignore. It is also easy to replace the New Tab page with a Blank Page (also right at the top of the Settings page) if you really don't want to see it, and there are extensions for other New Tab Page options.
Sure, making a bit of money by setting the default search to Google and showing a couple icons of brands isn't exactly "user-focused" but it's just business, and it doesn't seem to be getting in the way of prioritizing user features.
The MAC UI itself allows to assign a domain to always open in a specific container. If you need it more specific than an entire domain there are other extensions that can allow you to automate containers, but personally all I've needed are the domain-wide assignments.
Admittedly, I also tend to use unloaded tabs more than bookmarks, so if I need a very specific address in a specific container I am probably likely to just unload it but leave it open somewhere. (Sideberry "helps" me in this bad habit with multiple tab panels and tree grouping of tabs.)
For me the killer feature is tab stacking. It's crazy that no browser other than Presto-era Opera has this feature, that with workspaces is _so_ useful for tab organization.
I was looking at different chrome alternatives and most of them had ~something~ wrong with them. I have not tried vivaldi yet, but from a bit of research it seems just like the browser I was looking for. Thanks for sharing.
Any specific thing a first time user should know? Use build in ad blocker or install uBlock Origin?
This is false. You're welcome to install whatever air-conditioning system you want, the only requirement is that it needs to be deployed by a certified installer. Which makes sense, of course, we wouldn't want DIY setups that can say leak toxic chemicals or cause electrical fires all over the place.
As you said, you are not allowed to install it hence it is forbidden for you, you need a professional to do it for you. It is 400€ to buy a unit, but if you ask an professional shit cost 2500€ at least. r32 aint toxic and it is only mildly flammable and a few 100 Gramms, meanwhile anyone can use their natural gas cooker in their house alongside a 11kg bottle. I don't see the electrical hazard if you plug a small unit into a normal power socket. They teach how to wire a plug in school in the UK.
There is a big difference between mandating products offered to the public need to meet basic safety norms and mandating that you obtain the services of a third party cartel to maintain your own property.
You are not, in the UK at least they teach you how to install a plug yourself in school. Please tell me what is the hazard of adding a plug to your aircon?
It would be cool if Apple/Google/gatekeepers considered similar measures for the App Store / Google Play related search where similar constraints apply.
