In really grinds my gears that the buying companies take out the debt to take over against the companies themselves.
So many well-known UK companies have been sunk by debt interest on loans taken out to acquire said companies.
By all means use the companies to secure loans, but the liability should be on the books of the parent companies not the companies being acquired!
There have even been cases where the companies have been effectively asset-stripped by "sell and lease back" of property, leaving the companies a shell of their former selves with no meaningful assets, so as soon as there are any unexpected headwinds they collapse.
It's the bank's problem. The bank is supposed to determine whether it's likely to be repaid in full, and if not, don't issue the loan which blocks the sale.
It read to me to be entirely generated. The lack of details that people would normally mention tripped my spidey-sense. ( Who wouldn't name-check the restaurant in the opening paragraph? )
A double check, the author appearing to take up blogging in 2023, mostly about data science, with all the tell-tale signs of generated posts.
I don't use Claude Code. I use my own handwritten agent (formerly using Pi) and know every token that goes into it. There are zero memories to confuse it. The system prompt is 200 tokens and completely self consistent.
Plus I've found that the only time models go above 100k tokens anyway is when they've started looping at which point it's much better to go back anyway.
Anecdotally most models know their recall is terrible (or have been trained to act as such), that's why they constantly reread files before editing or while reasoning.
I read it as a models performance being random and observed differences in the opinions are the results of the overinterpretation of the random outcomes.
I think however that some people seem to be always lucky which indicates that it is not random but rather some fixed differences between people and their environments.
A models adherence to some configuration is a matter of probability. There might be some underlying pattern, but as far as I understand this is not documented and it may be even impossible to do so. So people are just trying stuff and sharing what appear to work. There is no causal link anywhere in this recommendations, and is just based on spurious correlations.
I think that's issue, rather than 60K being small.
Most of the actual edits/changes I request to codex are solved within 100-150K tokens, beyond 200K I'd definitively try to restart the session as soon as I could as all models are horrible once you get across ~20% of the total context size. And this is while working on +million LOC codebases.
Problem I guess is that there is no solid and concrete evidence of this (to me [and others seemingly] obvious) degradation, but should be easy to prove, yet no one has time to sit down and show it :)
But the likelihood of a model getting minor details wrong once you're above some magical threshold between 15-20%, seems to skyrocket, and I hit that issue sufficient amount of times that now my workflow is trying to prevent that.
what are y'all doing to hit that? Do you just not give it any pointers and let it churn away? What kind of context are you handing off?
I routinely get claude to do things pretty decently and finish up easily in the 4-5 digit range of tokens. It seems to be doing the right kind of thing to not waste its time looking at 1000 files.
And that includes London, it lists "excluding London" as £65k.
People overestimate how much senior devs in the UK earn, even after knowing they're not well paid, my usual response to hearing we should be earning £90k+ is, "well give us a job then"!
A friend is making about £180k / yr in London, and they bought a house recently in London. I think that's a lot, and his wife also makes a similar amount, slightly more. That seems to be the minimum, otherwise you're a renter for life. Pretty nuts.
A salary of £180k put you in the 98th percentile of UK salaries in 2024 (99th percentile was only a little higher at £207k) [1]. With a household income approximately doubling that, i'd suggest your friend is in an even smaller minority.
The average house price in London in July 2025 was £565k [pp33 - 2].
There is not being able to afford something and then there is not being able to afford exactly what you want.
Partly depends on what you mean by "London". I don't want to work for FAANG/AI/finance/gambling so our combined salary is well under the 180 figure. Yet we bought a 2-bed terraced house with a garden and driveway in south-west London (technically Surrey, but a London borough) just over five years ago. It's still 20 mins to Waterloo and most of central London is under an hour away. As we pay off the mortgage we should be able to further upsize over the years. It's not ideal, but it works.
That is an extremely high salary, and very far from the minimum required to buy, even on your own. A dual £350k income is truly astonishing. You could buy most houses in London in cash after saving for 5 years.
What? No that will not allow you to buy "most houses in London in cash after saving for 5 years" unless you live far out of town in a tiny place and eat plain rice for 5 years, and even then it'll be long odds.
First, you'll take home slightly over half of that net of various taxes and deductions, but let's be generous and say your take-home is 200k. You live very frugally, don't go out, don't really buy anything and keep your costs at 50k a year, including rent (!). That leaves you with 150k a year, so after 5 years you have 750k. This allows you to buy a modest 2-3 bed row house with a postage stamp sized garden in one of the less desirable areas of the city.
If you want something that doesn't look like a shed, you are looking at 1 million pounds and up, more like 1.5 million. If you want in a nice area and large garden, make it 2 million.
What are you smoking? The median house price in London is 500k. At 750k you can afford 77% of houses and at a million you are in the top 10%. 50k per year is also a preposterously high expenditure. Rent will be your leading expense by probably a factor of 10. You could put aside 3k a month for rent (again, totally preposterous number) and not touch the sides.
The only thing I can think of that would even come close to making a difference is having children. Then all bets are off, they can cost as much as you like.
But it sounds like it's not even a harness issue if they have a process where they send a reset email to an address that isn't associated with the account.
This isn't (just) a validation issue, and shouldn't be at the harness level.
I'm saving this comment, thank you for a great explanation of what it is like.
I was in my late 20s when I realised I was "face-blind", but I should have realised a lot earlier, I remember reading in a book as a child about how "people can recognise a person by their face from a long distance, but find it difficult to recognise a voice", and I could not relate whatsoever to that passage.
I thought I regularly struggle to recognise someone until they start speaking, but it wasn't until a decade or two later that I read about prosopagnosia and then suddenly a lot of things made sense.
Your explanation is so much better than the rubbish illustrations of blanked or blurred faces, because it isn't like that at all, indeed sometimes I might rely on a detail about their face to recognise someone.
It's why face-blind isn't a great term either, because it's not a kind of blindness, I can see just as well as anyone, it just doesn't trip the automatic and instinctual recognition that I understand most people have.
So many well-known UK companies have been sunk by debt interest on loans taken out to acquire said companies.
By all means use the companies to secure loans, but the liability should be on the books of the parent companies not the companies being acquired!
There have even been cases where the companies have been effectively asset-stripped by "sell and lease back" of property, leaving the companies a shell of their former selves with no meaningful assets, so as soon as there are any unexpected headwinds they collapse.
reply