| ▲ | davedx 2 days ago |
| The great thing about LLMs being more or less commoditized is switching is so easy. I use Claude Code via the VS Code extension. When I got a couple of 500 errors just now I simply copy pasted my last instructions into Codex and kept going. It's pretty rare that switching costs are THAT low in technology! |
|
| ▲ | shermantanktop 2 days ago | parent | next [-] |
| It’s not a moat, it’s a tiny groove on the sidewalk. I’ve experienced the same. Even guide markdown files that work well for one model or vendor will work reasonably well for the other. |
|
| ▲ | bgirard 2 days ago | parent | prev | next [-] |
| The switching cost is so low that I find it's easier and better value to have two $20/mo subscription from different providers than a $200/mo subscription with the frontier model of the month. Reliability and model diversity are a bonus. |
| |
|
| ▲ | paxys 2 days ago | parent | prev | next [-] |
| Which is exactly why these companies are now all focused on building products rather than (or alongside) improving their base models. Claude Code, Cowork, Gemini CLI/Antigravity, Codex - all proprietary and don't allow model swapping (or do with heavy restrictions). As models get more and more commoditized the idea is to enforce lock-in at the app level instead. |
| |
| ▲ | alecco 2 days ago | parent | next [-] | | FWIW, OpenAI Codex is open source and they help other open source projects like OpenCode to integrate their accounts (not just expensive API), unlike Anthropic who blocked it last month and force people to use their closed source CLI. | | |
| ▲ | gundmc 2 days ago | parent | next [-] | | Gemini CLI is open source too, though I think the consensus is it's a distant third behind Claude Code and Codex | |
| ▲ | _aavaa_ 2 days ago | parent | prev [-] | | The classic commoditize your complements. |
| |
| ▲ | bloppe 2 days ago | parent | prev [-] | | I only integrate with models via MCP. I highly encourage everybody to do the same to preserve the commodity status |
|
|
| ▲ | LetsGetTechnicl 2 days ago | parent | prev | next [-] |
| Using "low cost" and LLM's in the same sentence is kind of funny to me. |
|
| ▲ | harrisi 2 days ago | parent | prev | next [-] |
| I genuinely don't know how any of these companies can make extreme profit for this reason. If a company makes a significantly better model, shouldn't it be able to explain how it's better to any competitor? Google succeeded because it understood the web better than its competitors. I don't see how any of the players in this space could be so much better that they could take over the market. It seems like these companies will create commodities, which can be profitable, but also incredibly risky for early investors and don't make the profits that would be necessary to justify the evaluations of today. |
| |
| ▲ | crazygringo 2 days ago | parent [-] | | > If a company makes a significantly better model, shouldn't it be able to explain how it's better to any competitor? No. Not if it's not trained on any materials that reveal the secret sauce on why it's better. LLM's don't possess introspection into their own training process or architecture. | | |
| ▲ | harrisi 2 days ago | parent [-] | | That's my point. Anything that could exist that's significantly "better" would be able to share more about its creation. And anything that could be significantly better would have to be capable of "understanding" things it wasn't trained on. | | |
| ▲ | Smaug123 2 days ago | parent | next [-] | | Could you share anything about your creation, without having been to school where we taught you what the answers were? Can you deduce the existence of your hippocampus just by thinking really hard? | |
| ▲ | crazygringo 2 days ago | parent | prev [-] | | That's not true. There are a million ways to be "significantly better" that don't involve knowledge about the model's creation. It can be 10x or 100x or 1000x more accurate at coding, for example, without knowing a single thing more about its own internal training methodology. |
|
|
|
|
| ▲ | skydhash 2 days ago | parent | prev | next [-] |
| > It's pretty rare that switching costs are THAT low in technology! Look harder. Swapping usb devices (mouse,…) takes even less time. Switching wifi is also easy. Switching browser works the same. I can equally use vim/emacs/vscode/sublime/… for programming. |
| |
| ▲ | pchristensen 2 days ago | parent | next [-] | | Switching between vim <-> emacs <-> IDEs is way harder than swapping a USB (unless you already know how to use them). | | |
| ▲ | cozzyd 2 days ago | parent [-] | | I don't know, USB A takes 3 attempts to plug in for some reason. | | |
| |
| ▲ | ahmadyan 2 days ago | parent | prev | next [-] | | good point, they are standards, by definition society forced vendors to behave and play nice together. LLMs are not standards yet, and it is just pure bliss that english works fine across different LLMs for now.
Some labs are trying to push their own format and stop it. Specially around reasoning traces, e.g. codex removing reasoning traces between calls and gemini requiring reasoning history. So don't take this for granted. | | |
| ▲ | crazygringo 2 days ago | parent [-] | | I dunno. Text is a pretty good de facto standard. And they work in lots of languages, not just English. |
| |
| ▲ | jononor 2 days ago | parent | prev | next [-] | | You would have to buy said USB device and get it to your location.
Switching WiFi is only easy if you mean on a single machine/gateway. Swapping the WiFi network equipment in an office is considerably more involved, depending on the desired configuration. | |
| ▲ | amelius 2 days ago | parent | prev | next [-] | | You make it sound like lock-in doesn't exist. But your examples are cherry picked. And they're all standards anyway, their _purpose_ was for easy switching between implementations. | |
| ▲ | NicuCalcea 2 days ago | parent | prev | next [-] | | Most people only have one mouse or Wi-Fi network. If my Wi-Fi goes down, my only other option is to use a mobile hotspot, which is inferior in almost every way. | | |
| ▲ | oneeyedpigeon 2 days ago | parent [-] | | > Most people only have one mouse Tell me you're not a Mac user without telling me you're not a Mac user... | | |
| |
| ▲ | davedx 2 days ago | parent | prev | next [-] | | Huh? I have the VSCode extensions for both. Switching is a couple of mouse clicks and copy paste. | |
| ▲ | whatever1 2 days ago | parent | prev [-] | | I mean sublime died overnight when vscode showed up. |
|
|
| ▲ | falloutx 2 days ago | parent | prev | next [-] |
| on some agents you just switch the model and carry on. |
|
| ▲ | benterix 2 days ago | parent | prev [-] |
| Except Kimi Agent via website is hard to replace - I tried the same task in Claude Code, Codex, and Kimi Agent - the results for office tasks are incomparable. The versions from Anthropic and OpenAI are far behind. |
