I have used next in a pretty large application. Also built decently large things with base react (vite), vue (nuxt). Also small stuff using other platforms.

Next is by far the worst thing ever. Everything is broken in a weird way and the deeper you dig into it deeper it starts biting back. It was the only thing I dreaded maintaining.

Now that LLMs write code, it probably is not be as big of an issue though.

Still would not wish next upon any of my enemies.

Yeah. D1 is nice and easy but a serverless postgres offering would make a significant difference.

I was cursing at Codex for failing to push a PR. Trying a bunch of obscure methods one after another.. These tools should be trained to expect that GitHub going down is perfectly normal.

Same but w/ Claude Code. Was wondering why I was getting internal server errors when trying to push.

How does Zigbee compare to Matter? (as in, devices in the market, I know they are different levels but from a consumer perspective, there are zigbee switches and matter switches..) I've been using Matter switches because of HomeKit. Been pretty happy.

I do not have any Zigbee devices but I have many Hue devices. Their connectivity is shit. Button presses are not being registered. Sometimes commands do not reach devices etc. That's why I tried to avoid Zigbee.

But the low power battery operated zigbee stuff are pretty attractive if things work fine.

Hue is super-reliable for most people. Easily one of the best Zigbee implementations out there. Most likely, 2.4 GHz WiFi channels (or Thread channels) are overlapping with your Hue Zigbee channel.

https://www.oscium.com/training/zigbee-wifi-coexistence/

Deploying 2.4GHz Wifi, Zigbee, and Thread near each other requires some planning + corresponding configuration. A lot of Zigbee hubs try to find a quiet channel during the initial setup, but it does not always work.

How does Zigbee compare to Matter?

Matter over Thread is the most comparable to Zigbee (there is also Matter over WiFi). It's vastly more complex, since it uses IPv6 (ULE), mDNS, non-radio links between TBRs, provisioning is typically done through a smartphone with Bluetooth LE [1] etc. But it's also more standardized in that Matter defines the device types, so in principle a device will work with any Matter server, whereas with Zigbee support generally needs to be added per device (though there are generic drivers as well).

Big tech and IoT companies love Matter, because they can eventually gather analytics from devices or require cloud subscriptions. This is not just the case with Matter over WiFi, but Matter over Thread already allows devices to use NAT64 to connect to IPv4 servers on the internet (and TBRs like Apple TV do not allow you to disable NAT64). Thread 1.4 also allows for Thread devices to get globally routable IPv6 addresses. The Thread 1.4 white paper does not even try to hide what this could be used for:

Thread devices can now seamlessly connect directly to cloud services, enabling remote control, monitoring, and over-the-air firmware updates.

https://www.threadgroup.org/Portals/0/Documents/Thread_1.4_F...

So, yeah, Matter + Thread does some things better, some things worse, and opens up new avenues for monetization that were not possible with Zigbee, because Zigbee devices were fully local and owned.

[1] There are workarounds, e.g. Home Assistant allows provisioning through Bluetooth proxies. Also, provisioning does not currently work on alternative ROMs like GrapheneOS.

That's my main problem with Thread. With Zigbee I have confidence that the mesh is 100% local and offline. How do I get that confidence with Thread? How can I really be sure that something isn't helpfully opening it up to the world?

I have no problems with zigbee. Here it is 33 zigbee devices, 16 being routers (light/power plug switches). No basement.

My friend has problems with zigbee and has moved to matter and is now trouble free. The only fine zigbee stuff for him was the Hue lights though.

I have seen repeated discussions in Home Assistant adjacent forums that Matter based devices have considerably shorter battery lives, compared to zigbee things that run on one CR2477 cell. Or things that use zwave800 chipsets. I have no Matter devices myself, but I've seen it often enough and from different people who don't seem clueless that it is concerning.

Isn't that inevitable?

I used to keep an eye on the web frameworks. Tried Solid JS, Vue, Svelte etc regularly to see what's going on. But at this point, if I'm not gonna enjoy the platform or suffer from it then it might as well be jquery + imperative code.

Also, web frameworks are not really about coding anyway. These are HTML generators. They were about developer ergonomics beyond anything and were important when it was about developers.

I am subscribed to ChatGPT, Claude, Kimi and GLM coding plans. 200$ one on GPT and the 20$ ish ones on all others. Recently added Grok and it has somehow bacome my second most used model.

For daily one off questions I prefer it because it is fast enough and I like the way it responds. I also use it for basic research like “find me a battery drill for this and that”.

Kimi and GLM feel extremely coding oriented. I use them for code reviews basically. I hate the way Anthropic models talk. GPT takes too much time and effort for that kind of stuff for some reason.

Grok happened to be a nice middle ground.

It is still not multi language. I mean I can talk to ChatGPT, Claude, Gemini, fucking Grok in any language I want and it will understand / respond in given language. Apple being Apple, I have to choose a language in Siri menu and it has to be the phone language for some reason.. Otherwise it downgrades to old Siri.

Apple's incompetence in AI / Assistant etc stuff has to be studied.

What is amazing to me is that when driving, Siri used to read out messages sent to me in other languages in the original language.

Now it just says something like "A message in a different language".

I think Apple's top-down approach is showing here. Google is more bottom-up, which has its flaws, but it's more suited for cases like this one. I seems that unless one of Apple's higher-ups notices a problem nobody is going to do anything.

For example, Android and Google Assistant has had excellent support for Polish for more than a decade. Siri still doesn't support Polish at all, even though all of Apple's AI suppliers support Polish perfectly well...

> Apple's incompetence in AI / Assistant etc stuff has to be studied.

Why should Apple focus on technical excellence when the company and their products are touted for literally gimmicks? In the end this is what brings the money and profits through the door, can't blame a company for providing what the market craves.

The market craves good AI experiences... But that doesn't mean Apple is capable of delivering it at this stage.

Does the market crave good AI experiences?

LLMs are eating software, maths, and LinkedIn posts; some of the image generators are seeing similar (limited) success.

Other than that, my general impression is that most people think of even the best as garbage.

("Other than". In fairness, most people thought LinkedIn posts were already garbage, it's just that it wasn't made any worse by LLMs).

This post is a testament that mobile AI experience on Android is much better and yet iphones are not really losing market share to that.

I wouldn’t use an immortality serum if it came with Norton branding. How come it is still a thing?

> which they clearly think is bad

These are companies with huge amounts of money invested in them. Their main responsibility is to their investors. If building better models (or making more money from the models they already have) would increase returns, they would do it. So they clearly can’t do it right now.

Climate change hasn’t stopped the oil industry, and concerns about forever chemicals haven’t stopped the chemical industry. Companies don’t usually give up profitable opportunities just because they might be harmful or kill us all in the long term..

I think you fundamentally misunderstand what kind of company Anthropic is. I'm simply taking the path of least resistance when understanding what people at Anthropic have to say, and basically every public source points to Anthropic talking the talk as well as walking the walk with regards to AI safety and their ideas on AI governance.

Re the point on returning shareholder value, Anthropic is a PBC:

> Public Benefit Purpose: Under Delaware law, a PBC allows directors to legally weigh social and public good alongside maximizing profit for stockholders

Investors surely know this and Anthropic probably has pretty reasonable legal grounds to stand on if they chose to prioritize safety over profits.

Seems pretty clear to me that this is in earnest.

The decision to found Anthropic to compete with OpenAI's then concentration of SOTA LLM AI researchers, and then achieving/surpassing parity with OpenAI is pretty much directly responsible for creating the current capabilities arms race.

Google has shown they are not motivated at all to push the bigger and bigger SOTA frontier with Mythos-type super models. Even OpenAI isn't really even trying to surpass Anthropic at the top of the market, preferring to keep pace in benchmarks but at lower costs to serve. Anthropic with Mythos is now more or less alone in aggressively pushing the ceiling for the most "super-intelligent" models current computing can support.

It has made them and their investors a gigantic pile of money, but we're really supposed to believe their contribution to that arms race (which they now describe as a threat to humanity) has always been completely altruistic and that their corporate structure inherently makes self-interested decision making impossible?

If you've read enough lesswrong, this won't be new to you. The strategy always was "become the most capable model lab so that you have leverage to control AI development."

Oh I am terribly sorry, did not know Anthropic was formed as a a “Good Company”. Which changes everything ever.

Snark aside, maybe I’m too pessimistic and you are correct on this. But.. doesn’t that mean we are completely fucked? If they don’t have any ulterior motive and they truly are panicking.. We should start packing.

yes that is literally why they are sounding the alarm - because people think AI is a party trick and they aren't taking it seriously enough.