olliestanley
olliestanley.bsky.social
olliestanley
@olliestanley.bsky.social
ML engineer. Increasingly infrequent forays into armchair econ/policy
There has never been a chancellor who operated as incompetently as Reeves, though, barring perhaps Kwarteng who got a far worse hammering than Reeves
November 30, 2025 at 9:08 PM
Remember the median Labour member believed (something ~as stupid as) "there was always plenty of money but the Tories were siphoning it off to their mates rather than funding services"

Was always going to be a rude awakening realising the services are intrinsically unaffordable regardless of govt
November 9, 2025 at 11:14 AM
Yeah, I tried to use this site for a while, and it does offer a useful balance to the doom prophecies I'm subjected to on twitter, but discourse on here (vs twitter with a bit of effort to curate/maintain feed quality) is just garbage. I only come back once every few weeks now
November 2, 2025 at 6:01 PM
This is mostly just about the after tax wages smart young people can get. Those started dropping rapidly around 2000, and there was a lag as the info filtered through. Brexit made it worse but 80% of this was locked in beforehand
November 2, 2025 at 5:39 PM
IE's views are already about as centrist (if anything slightly left of centre) as it's possible to get in the context of the wider electorate, a "centrist dad" on bluesky more or less just means a socialist anywhere else
November 2, 2025 at 5:37 PM
Circular argument. Politicians do this because they are incentivised to by the electorate's behaviour. Any politician who doesn't is the lone cooperator in a prisoner's dilemma and will simply lose to those who defect. Hard truths don't work
November 2, 2025 at 5:29 PM
I mean, you are on a website that's almost exclusively for socialists & communists. So it makes sense
November 2, 2025 at 5:26 PM
Was there more depth to the point? As written here it seems to be a reiteration of something people have been saying on here/X for at least 3-4 years
September 29, 2025 at 10:11 AM
Odd to suggest this is "private equity" when it's just the costs of premises & labour. Hence you can still get good cheap food, even in London, at restaurants that have been around ages and are run by owners, since they own premises outright (no rent/mortgage) & save on labour by doing it themselves
September 29, 2025 at 10:09 AM
The point of them being on X is to reach out to and convince more people. A large proportion of the voting population have X accounts but don't and never will have BSky accounts. Abandoning X just means it, and British politics as a whole, gets more right wing
September 14, 2025 at 8:43 PM
Not a bad prediction imo. Would restore British politics to something more historically normal & in line with most of Europe
July 26, 2025 at 9:30 PM
Great. Then they save more of the money they would've spent on luxuries, slowing the economy and increasing "wealth inequality" further.
July 26, 2025 at 9:14 AM
Without a platform driver ideological thumb any recommendation algorithm would just tend towards, and amplify the ideological bias of the users, meaning the feed would be full of extreme communism. Strongly suspect this is why the current Discover is so neutered - they want to avoid that outcome
July 9, 2025 at 9:33 AM
People also rate "decisions made by Starmer and Reeves" as the leading factor in the weakness of the economy so seems they believe October's tax rises have been bad for economic growth www.theguardian.com/politics/202...
July 8, 2025 at 8:11 PM
Meanwhile, Labour polling is flat since they announced the welfare changes, and has risen since they started trying to get them through Parliament yougov.co.uk/topics/polit...

They overall seem to have had little to no impact with the wider public -> tax rises seem more toxic than spending cuts
July 8, 2025 at 8:11 PM
Evidence is that the October budget was a major problem (see for e.g. www.moreincommon.org.uk/latest-insig...)

Though winter fuel was a part of that as well as tax, anecdotally seems tax rises have remained toxic in public discourse ever since, while winter fuel faded until the high-profile U-turn
July 8, 2025 at 8:11 PM
Even a relatively small NI rise has been unbelievably toxic for Labour. From polls tax rises seem the biggest thing that have damaged Labour with the public, compared to their proposed spending cuts which were of course highly toxic with small interest groups but lacked real widespread poll impact
July 8, 2025 at 7:25 PM
In the long run it can be achieved yes, with broad based tax rises. Shorter term no. Our tax base is narrow, and the necessary broad-based increases would kill a government at the next election.
July 8, 2025 at 7:22 PM
Large tax rises seem no more politically sustainable than large spending cuts, to be honest. Between demographics & productivity we've already crossed the point of no return; restoring fiscal sustainability will necessarily be politically toxic. Only a suicide one-term government can fix the UK
July 8, 2025 at 6:47 PM
Sadly though, a lot of people did say that. Many prominent Labour politicians constantly argued in opposition that there was plenty of money and the only problem was the Tories giving it to their mates / cutting taxes for billionaires / etc
July 8, 2025 at 10:49 AM
Non-dom and school VAT already widely documented, but here's the early evidence from OBR forecasts that the CGT raise is actually costing revenue, not raising it

Here some data suggesting Labour's stamp duty raise is going to be a net cost to the Treasury too: www.telegraph.co.uk/money/proper...
July 7, 2025 at 12:28 PM
As a medium-large company I think it would be a poor strategy to vendor lock yourself in with a model provider which will never offer top tier capabilities just because it can still do some of the routine tasks. Competitors will surely accept higher costs in order to automate more complex tasks
July 7, 2025 at 7:47 AM
For me Grok and DeepSeek show that though many labs can produce a model at 80% of frontier quality, we have still never seen anyone (arguably Anthropic briefly) catch OpenAI / DeepMind, and power/commercial users are willing to pay a premium to have the best model over a model just nearly as good
July 7, 2025 at 7:14 AM
Not convinced tbh. Smaller, more targeted companies (Cursor being the most obvious) are beating frontier labs at building products targeting specific use cases, so it seems that selling general-purpose API inference will be critical for the frontier labs
July 7, 2025 at 6:53 AM
Agree R&D costs will remain high to keep them ahead, but since inference itself seems to be profitable (and becomes more so as models become more useful + inference stack is optimised) it seems plausible that inference profits can ultimately exceed R&D costs
July 6, 2025 at 8:15 PM