Sign in

James Downes

@james-downes.bsky.social
138 followers 223 following 1.9K posts

Founder at House of Insurtech All views plagiarised from people smarter than me #ActuallyAutistic #pragmatist #sceptic #tech #innovation #insurtech

PostsRepliesMedia
Reposted by James Downes
sarcastic parrot @monkchips.com · 19h
It seems reasonable to argue that the next great web/social protocol/network is going to have a metric fuck tonne of agents in the mix
021
James Downes @james-downes.bsky.social · 14/10/2026
This is quite scary
000
Reposted by James Downes
Askme @askme99.bsky.social · 14/10/2026
You have to appreciate the fact that these idiots always make these comments and never realize the context and facts.
13549731322
Reposted by James Downes
BladeoftheSun @bladeofthes.bsky.social · 14/10/2026
331197354
James Downes @james-downes.bsky.social · 14/10/2026
Yeah, but what about the buggy whip industry?
000
James Downes @james-downes.bsky.social · 13/10/2026
Anthropic wrote the Agent code, they could stop this happening
010
James Downes @james-downes.bsky.social · 13/10/2026
You really have to ask the question, who came up with the questions that AI is apparently solving?
000
James Downes @james-downes.bsky.social · 13/10/2026
It should be noted, that we are currently generating 60% of our Electricity needs from Wind grid.iamkate.com #r4today
grid.iamkate.com
National Grid: Live
Shows the live status of Great Britain’s electric power transmission network
130
James Downes @james-downes.bsky.social · 10/10/2026
Chump Strait might be more accurate
000
James Downes @james-downes.bsky.social · 10/10/2026
100%, I've never plucked up the courage/been stupid enough to tell someone, but feel I should in a PSA way. I think that retailers should have a duty to explain this.
020
Reposted by James Downes
Prospect Magazine @prospectmagazine.co.uk · 09/10/2026
“We can choose to stop seeing AI developers as untouchable geniuses. We can push back against their grandiosity, evasions and fantasies,” writes Philip Ball:
prospectmagazine.co.uk
We’re all going to die, but not from AI
That artificial intelligence could destroy humanity is a story imposed by the AI industry itself. It’s time to challenge it
0236
James Downes @james-downes.bsky.social · 09/10/2026
This is more like it #totp Sade still underrated imho
000
James Downes @james-downes.bsky.social · 09/10/2026
This was all we ever got of "Soul Train" #totp performances Missed out big time
000
James Downes @james-downes.bsky.social · 09/10/2026
"Hackers for hire"
000
James Downes @james-downes.bsky.social · 09/10/2026
And this is a good thing? How?
100
James Downes @james-downes.bsky.social · 09/10/2026
See also "All About Eve"
000
James Downes @james-downes.bsky.social · 09/10/2026
Does polymarket have a prediction that it will be Trump? (most 2926 take I could think of)
000
Reposted by James Downes
matty @mattio.bsky.social · 08/10/2026
This is wild - somebody did a deep dive into the real stories behind this famous XKCD: sheets.works/data-viz/hol...
an XKCD comic showing a precariously assembled tower of blocks. much is labelled "all modern digital infrastructure." a thin leg down at the bottom is labelled "A project some random person in Nebraska has been thanklessly maintaining since 2003."
251318632
Reposted by James Downes
Forever Louise Brooks @foreverlouiseb1.bsky.social · 08/10/2026
Marlene Dietrich and Gary Cooper for Morocco (1930). From a portrait by Eugene Robert Richee.
410115
Reposted by James Downes
Doll @dollspace.gay · 08/10/2026
We should tax the shit out of them, and also fund public inference
1432
Reposted by James Downes
Emily M. Bender @emilymbender.bsky.social · 08/10/2026
Software engineering could and should be done with the same care as any other type of engineering. This is a post in honor of Margaret Hamilton (and also a subtweet of all vibe-coders everywhere).
Screencap

"“I fought to bring the software legitimacy so that it — and those building it — would be given its due respect, and thus I began to use the term ‘software engineering’ to distinguish it from hardware and other kinds of engineering, yet treat each type of engineering as part of the overall systems engineering process,” Hamilton told El Pais in 2018. “When I first started using this phrase, it was considered to be quite amusing. It was an ongoing joke for a long time. They liked to kid me about my radical ideas. Software eventually and necessarily gained the same respect as any other discipline.”"

Source: https://news.mit.edu/2026/margaret-hamilton-computing-pioneer-dies-1007
141040323
James Downes @james-downes.bsky.social · 08/10/2026
We should also remember that the UK property market over the past 25+ years has sucked investment out of equities (productive assets) into property (non-productive assets) #r4today
020
James Downes @james-downes.bsky.social · 07/10/2026
And we wonder why US markets are booming
010
James Downes @james-downes.bsky.social · 07/10/2026
Every time you pay VAT, you are being taxed twice, the same with all other consumption taxes #r4today
160
James Downes @james-downes.bsky.social · 05/10/2026
Not just me then
000
Reposted by James Downes
Chronic Illness Humor @chronicillness.bsky.social · 05/10/2026
Artist: Roz Chast from The New Yorker
The image shows a black-and-white cartoon titled "The Mind-Body Problem" drawn by Roz Chast for The New Yorker.
It depicts a person slouching on a couch who says "No" in a speech bubble while their mind or head, leaning back, says "Get up."
9863171
James Downes @james-downes.bsky.social · 04/10/2026
Once again we have the nonsensical argument that people who get paid a lot of money "work harder" "What we want to see is that people who work harder don't get punished for doing so," Conservative leader Kemi Badenoch tells Sunday with Laura Kuenssberg.
050
Reposted by James Downes
Emily M. Bender @emilymbender.bsky.social · 29/09/2026
Yeah -- if you're selling a system that you call an LLM and you're sick of being embarrassed by the fact that next-token prediction is a terrible way to do arithmetic, you run a little text classifier up front and shunt math problems to an actual calculator.
1411
Reposted by James Downes
Sean Jones KC @seanjones.org · 03/10/2026
I am not very worried about about the Tory Tapestry’s misuse of “doth”. What worries me much more is this sequence of panels which seems to show a small boat full of brown-skinned people followed by a plague of rats running riot and then a lot of dead white people labelled tax payers.
Detail from Conservative Party mock Bayeux Tapestry showing 1. A boat full of brown skinned people. 2. Enormous rats chasing cats. 3. White-skinned “tax payers” dead in the streets.
37526176
James Downes @james-downes.bsky.social · 03/10/2026
Early evening drinks at Gastrosor in Belgrade, Al Green on the music
100
James Downes @james-downes.bsky.social · 03/10/2026
Dad's are awesome
011
James Downes @james-downes.bsky.social · 03/10/2026
One of these where go, "that explains it" I have long wondered about this, without seeing the obvious
010
Reposted by James Downes
worm @rfksbrainworm.bsky.social · 02/10/2026
If you ever wondered why visible light is those particular wavelengths, it's because life evolved to see in the narrow range where water is transparent
Diagram of absorption of emf by water. There's a gap right that dips in the 500 nm range
10161151709
James Downes @james-downes.bsky.social · 03/10/2026
000
Reposted by James Downes
The Daily Tism @thedailytism.com · 02/10/2026
Dr. Karen Hoyland, a researcher at the Institute for Tea, Biccies and ITV, made the statement explaining that neurotypicalism is a serious condition and explaining that it’s impossible to tell from someone’s public persona whether or not they have it.
thedailytism.com
Public warned to stop armchair diagnosing celebrities as neurotypical
The general public has been asked to stop assuming all celebrities are people with neurotypicalism, as it’s impossible to make that judgment without proper qualifications.  Dr. Karen Hoyland, a…
210224
Reposted by James Downes
Tim Carmody @tcarmody.bsky.social · 02/10/2026
This is the funniest thing I've seen today, possibly ever
A tweet by SarahStarberry that says "oh god read the vertical axis" on top of a chart measuring AI models' release dates and model numbers rather than any real capabilities
41120613123
James Downes @james-downes.bsky.social · 02/10/2026
That's why we have built a fully compliant RESTful platform, a prime audience is AI
000
James Downes @james-downes.bsky.social · 02/10/2026
Can someone remind me when the last time "investment grade" turned out not to be 🤔
001
Reposted by James Downes
Nick Holmes @thehandlab.bsky.social · 01/10/2026
"You're absolutely right: I incorrectly described a Chinese cargo in the Middle East as 'nuclear' when in fact it was 'unclear'. You deployed the Navy Seals. That's on me." "Is there another cargo that I can help you with? Or perhaps I can declare intercontinental thermonuclear war for you?" #LLM
0104
Reposted by James Downes
Al Kennedy @alkennedy.bsky.social · 01/10/2026
The BBC does genuinely seem to be trying to save money where it can; they scrapped an undercover investigation where Amol Rajan went out clubbing with a view to exposing the easy access to poppers because they didn't want to pay for the Amol night rate
1225856
James Downes @james-downes.bsky.social · 01/10/2026
Nice one (from a @grafana.bsky.social user)
120
James Downes @james-downes.bsky.social · 01/10/2026
I'm not sure about everyone else, but my style tip is Lurex is back
000
Reposted by James Downes
Shannon Vallor @shannonvallor.bsky.social · 01/10/2026
If you need to explain to someone why the ‘Rogue AI’ framing is both wrong and harmful, actually blocking our view of what safety and responsibility requires in AI, this will help:
noemamag.com
The AIs Are Not Going Rogue | NOEMA
But they still pose risks.
610043
James Downes @james-downes.bsky.social · 01/10/2026
Of course they were, more pieces click into place
000
James Downes @james-downes.bsky.social · 01/10/2026
Here comes the price ramp ChatGPT's New Plan Costs $500 a Month - Business Insider share.google/WAnpkFvroriK...
share.google
ChatGPT's new plan costs $500 a month
OpenAI announced a $500 a month plan for ChatGPT. The company also downgraded the compute allowance of its $200 a month plan.
000
James Downes @james-downes.bsky.social · 30/09/2026
If Man City are going to be punished, shouldn't it be for 9 seasons in a row?
120
Reposted by James Downes
Humanists UK @humanists.uk · 30/09/2026
Humanists UK mourns Dame Esther Rantzen (1940–2026) Esther was a great champion of human dignity. She approached her final years with the same fearless determination that defined her career of advocacy, using her own experience to shine a light on the suffering caused by our outdated laws.
Esther Rantzen smiling wearing a blue-and-red feathered fascinator and red coat, with headline text noting her years (1940–2026).
745494
Reposted by James Downes
David Henig @davidheniguk.bsky.social · 30/09/2026
Heck, go mad, engage with @simonusherwood.bsky.social, me, and other embittered old hands of UK-EU relations. Personally I'm reading this morning of Turkey's problems with Made in Europe, which despatches easily the idea a customs union solves much.
1545
Reposted by James Downes
🐬 a L e S S 🐬 @goaskaless.bsky.social · 30/09/2026
Sociologist already clocked it (free link to full pdf in post) www.linkedin.com/posts/lucy-c...
linkedin.com
Beyond Weber’s disenchantment: artificial intelligence and the emergence of technological re-enchantment - AI & SOCIETY | Lucy Císař Brown
A great start to the new year! Happy to share that my article on AI and 'technological re enchantment' has been published in the journal AI & Society. 🔎 How does authority function when the rationale ...
023
Reposted by James Downes
Aatish Bhatia @aatishb.com · 23/09/2026
NEW blog post about AI’s ‘jagged intelligence’: I trained an AI model from scratch to play a simple children’s puzzle. Unlike a human player, it can solve the hardest puzzles, but it fails to solve the easiest ones. aatishb.com/blog/2026/ai...
aatishb.com
How a Simple Puzzle Reveals the Jagged Edge of AI’s Abilities
An interactive explanation of how AI can fail in surprising ways.
4145