Because of extra services. Unlike emails, which usually come with no guarantees or support, the money-sending services are protected. Even with something like EFT, consumers have 60 days in case of actual fraud or bank mistakes, and with credit cards, you are also protected from things like product non-delivery.
Those protections exist and cost money - both to pay for customer support handing those, as well as to reimburse customer when the original money cannot be found anymore.
(And related, this is one of the big problems with cryptocurrencies: most people are not happy to know that one typo / software bug means their money is lost forever. Those protections on regular baking are there for a reason.)
Gleam is my new default language and ecosystem. Until last year, I first reached for .NET on new projects. ".NET everywhere" and such. But Blazor apps are huge and slow. These days I often advocate first for Gleam. With Lustre and Tauri it can run anywhere.
Agree. A whole tree of comments agreeing with you is not appearing on this post because that's how HN works.
I loathe LLM prose, no matter who posts it. It earns a quick cmd+w.
My prizes for spending the internet points to say this are the joy of resisting normalization of illiteracy and a tacit camaraderie with the hundreds of others who agree but won't comment or upvote.
Experienced distance runner here. This is not new or mysterious.
> Overtraining Syndrome is where a lack of recovery from training stress results in underperformance, often combined with Mood Disturbances such as reduced Motivation, depression and anxiety
I am also worried about "cognitive debt". I hardly remember what I had Claude do, even hours later. Back in May, I advised of a similar mitigation, citing the "generation effect" as the reason typing the code would make you remember it better:
> For your next ticket or feature, engage with your LLM as you normally would to produce a design and implementation plan, but with that plan in hand, make the mechanical edits yourself...You can expect this to improve your knowledge retention compared to merely reading a diff. You want to go slower now so that later you can go at all.
the punchline: "Refactoring reduces token consumption"
I appreciate the effort to quantify the benefit rather than pontificate. It's worth mentioning Martin Fowler wrote a whole book on refactoring [1], in which he states, "to refactor, the essential precondition is [...] solid tests", which I think is the real benefit here, AI or not. Good tests protect against regressions, whether human or robot. They also help encode the spec, which humans and robots can read.
I think the real punchline was that the value saved was on the order of cents!
> Every single change that touches the data access layer from this point forward now costs significantly less.
> How much of a saving? Assuming Sonnet 5 pricing at the time of writing of $3/MTok, 39.7 cents.
Now consider OpenAI's price drop, and open models, and consider that in the long run tokens will get cheaper. And think that the refactor needs to be human guided at a price of what for a senior developer - $100/hour?
Only person asking this, but it's possible they ran out a statute of limitations by going down the wrong path. (SoL normally begins when the prosecution becomes aware that the crime has happened. In the USA SoL for most felonies is often 3 years.)