dan

Support is very widespread so clients that don't support it are likely using outdated libraries.

Yeah it's very common for CDNs to transcode images (convert them to a different format) to save bandwidth. Your browser or app tells the server which image formats it supports using the Accept HTTP header, and the server returns the image in the best format it supports. There's been significant improvements to data compression technology in the last 20+ years, and an AVIF file is usually around 50% the size of a comparable JPEG file of similar quality.

Cloudflare does this (although I'm not sure if it's only a paid feature or if it's available on free accounts too), and major services like Google, Facebook, Instagram, etc do it too. At those scales, halving the amount of bandwidth used by images is a significant saving.

If an app gets an image format it doesn't support, it's likely not setting the Accept header correctly.

Edit: There's also JPEG XL which is essentially supposed to be like "JPEG 2.0", but it's not widely supported yet.

In the 90s, my mum's washing machine would stop if you needed to rebalance it, and a "rebalance" light would come on to let you know. It doesn't seem like newer washing machines still do that?

This is interesting because even though they designed the language to be impossible for LLMs to use, Claude still scored nearly 70% on their benchmark.

Help someone in need.

AI is more than just chatbots though, and there's plenty of use cases where it can be used to help people rather than replace them.

I started the year being very sceptical of AI, but it's helping me with a bunch of things at work now. For example, at work we have an agent for data analysis that can find the correct tables to use, gather data, and produce a report, along with proof (all the SQL queries it ran plus any ad-hoc code it used for analyzing the data) so that the work is repeatable and we can verify the accuracy of the queries and logic.

I'm a developer and it's helped a lot with side tasks too. If I'm coding and see an issue in our code (usage of deprecated code, tech debt we need to pay down, etc), I used to file a task or add a todo comment to look into it later, and eventually get to it. A lot of them are small tasks, it's just that I don't always have time to look into them while in the middle of a project.

These days, instead of adding it to a todo list, I just spin up an agent and get it to look at it while I'm working on other things. The resulting code doesn't get submitted for code review until I manually look at it and ensure it's correct. The codebase I work on has never been cleaner, because AI has helped me clean up so much legacy cruft that's accumulated over the years, while I still deliver my regular projects on time.

My personal account cost 3 times as much for the same prompt on my work account.

That's interesting. My employer is too large to use subscriptions and also we have our own harnesses, so they have to pay per token. Some employees spend thousands of dollars (or more) per month, but if they were doing it at home, they'd be easily able to fit the same usage into a $200/m subscription.

I don't have data to back this up but I'd say thay the majority of their customers aren't also selling AI services themselves. It's customers that use AI for their business processes, software development, etc.

AI is an enterprise product

I think Google and Meta are well-positioned to provide AI services to regular people, since they both already have a large user base, existing cash flow from other non-AI products (meaning they can subsidize AI features for as long as they need to), and can probably outlast any startup. Google especially easily have enough paid users and advertisers to cover costs for free users.

Most of thr other AI companies will probably all end up focusing on / pivoting towards enterprise use cases though, for the higher prices and stability of long-term contracts.

AI predicts the next token

LLMs specifically... Not all AI does this.

Even so, modern LLMs are more sophisticated and aren't just chatbots any more. Their harnesses (the app that uses the LLM, like Claude Code, OpenCode, OpenAI mobile app, etc) let them use tools - read and write files, browse websites, hit APIs, run commands, etc. That's what makes them dangerous. Not all tools have good guardrails, and some "AI-native" people seem to like running harnesses in dangerous/YOLO mode.

Anthropic made a smart business decision to mostly focus on enterprise customers rather than regular consumers. Only around 10% of their revenue comes from their heavily subsidized subscription plans. The remaining 90%ish comes from API and enterprise usage billed at per-token rates, which is something that's actually profitable for them. They have a decent margin on token rates. It's R&D, hardware, and subscriptions that are putting them in the red at the moment.

I suspect they'll discontinue subscriptions some time after their IPO. Even though subscriptions are only responsible for 10% of revenue, they consume 42% of their inference compute power (https://www.theregister.com/ai-and-ml/2026/10/06/anthropic-claude-subscription-plan-provides-more-value-than-openais-study-says/5301470)

In comparison, around 70% of OpenAI's revenue comes from subscriptions. Without subsidized subscriptions, they'd lose most of their customers.

legacy tlds (com, net, org) tend to be more affordable

Verisign are trying their best to make .com and .net less affordable. Their contract with ICANN is on a six year cycle, and it lets them increase the price by up to 7% for the last four years. Two years with no increases, then four years with increases. So far, they've always increased it the maximum amount.

Wholesale price has gone up from $6 in 2000 to $7.85 in 2020 (before this contract structure), to $10.26 now, and they're increasing it to $10.97 on November 1.

Having said that, it's still better than ccTLDs and newTLDs, where the registry can increase it as much as they want.