Jon Ayre
@jonayre.uk
I've been an engineer, software developer, architect, director & CTO. Now I'm a hands-on consultant doing business and tech strategy. Creator of the business evolution map. Father & Husband. It WILL go wrong & it CAN be fixed. He/him.
I'm going to write a blog post about an idea I've had, but right now I'm on holiday so it'll have to wait. The concept is this: intelligence is difficult to define but easy to detect.
Intelligence reverses entropy.
Intelligence reverses entropy.
September 17, 2026 at 10:03 AM
I'm going to write a blog post about an idea I've had, but right now I'm on holiday so it'll have to wait. The concept is this: intelligence is difficult to define but easy to detect.
Intelligence reverses entropy.
Intelligence reverses entropy.
My opinions of the so called "AI threat" and calls by the main players for regulation haven't changed even slightly since this post 8 months ago.
Lobbying for regulation from a position of strength is a classic counter-mobility tactic.
Lobbying for regulation from a position of strength is a classic counter-mobility tactic.
My opinions of the so called "AI threat" and calls by the main players for regulation haven't changed even slightly since this post 8 months ago.
Lobbying for regulation from a position...
My opinions of the so called "AI threat" and calls by the main players for regulation haven't changed even slightly since this post 8 months ago.
Lobbying for regulation from a position of strength i...
lnkd.in
September 15, 2026 at 2:43 PM
My opinions of the so called "AI threat" and calls by the main players for regulation haven't changed even slightly since this post 8 months ago.
Lobbying for regulation from a position of strength is a classic counter-mobility tactic.
Lobbying for regulation from a position of strength is a classic counter-mobility tactic.
I love the Asimov robot stories. TBH I pretty much grew up on them, but anyone who thinks the three laws are in any way implementable needs to try coming up with an unambiguous definition for "harm".
Even Asimov made this point in his later stories.
Even Asimov made this point in his later stories.
September 13, 2026 at 1:18 PM
I love the Asimov robot stories. TBH I pretty much grew up on them, but anyone who thinks the three laws are in any way implementable needs to try coming up with an unambiguous definition for "harm".
Even Asimov made this point in his later stories.
Even Asimov made this point in his later stories.
Reposted by Jon Ayre
The theory's almost complete. Set your bar for shippable code to even just one-9's reliability, and the human has to stay very much in the loop.
Set it to two 9's, and the human *is* the loop.
codemanship.wordpress.com/2026/09/12/o...
Set it to two 9's, and the human *is* the loop.
codemanship.wordpress.com/2026/09/12/o...
Out-Of-Distribution Checks – How Human Intelligence Stabilises Agentic Workflows
Back to my “Special Theory of Autonomous Agent Reliability” (STAAR). The reliability of a step (a model interaction) in an agentic workflow: R = 1 – (1 – C)(1 – P) Whe…
codemanship.wordpress.com
September 12, 2026 at 7:18 AM
The theory's almost complete. Set your bar for shippable code to even just one-9's reliability, and the human has to stay very much in the loop.
Set it to two 9's, and the human *is* the loop.
codemanship.wordpress.com/2026/09/12/o...
Set it to two 9's, and the human *is* the loop.
codemanship.wordpress.com/2026/09/12/o...
"AI went rogue".
Nope. Hiding behind the machine was ways going to be the strategy:
Nope. Hiding behind the machine was ways going to be the strategy:
jonayre.uk
September 13, 2026 at 11:32 AM
"AI went rogue".
Nope. Hiding behind the machine was ways going to be the strategy:
Nope. Hiding behind the machine was ways going to be the strategy:
You know what would be really funny? If they still lost, even with the £72m and all they did was alert the taxman to the much larger crypto gains they converted to fiat currency.
September 13, 2026 at 9:21 AM
You know what would be really funny? If they still lost, even with the £72m and all they did was alert the taxman to the much larger crypto gains they converted to fiat currency.
A post in which developers compare their own code with that of AI & find AI writes better code than them. "Look how bad my old code is" they declare. It's fine to compare, but don't assume what you find applies to others. Maybe the AI is great at coding? Maybe you're bad at it? Either could be true.
A thread in which developers compare their own code with that of AI and conclude AI writes better code than them. "Look how bad my old code was" they declare. It's fine to compare, but d...
A thread in which developers compare their own code with that of AI and conclude AI writes better code than them. "Look how bad my old code was" they declare. It's fine to compare, but don't assume th...
lnkd.in
September 13, 2026 at 8:34 AM
A post in which developers compare their own code with that of AI & find AI writes better code than them. "Look how bad my old code is" they declare. It's fine to compare, but don't assume what you find applies to others. Maybe the AI is great at coding? Maybe you're bad at it? Either could be true.
I see they're still trying to make atproto a thing...
September 12, 2026 at 7:36 AM
I see they're still trying to make atproto a thing...
"I uninstalled my IDE now that agents exist".
Why?
Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
Why?
Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
"I uninstalled my IDE now that agents exist". Why? Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
...
"I uninstalled my IDE now that agents exist". Why? Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
By the time y...
lnkd.in
September 11, 2026 at 7:32 AM
"I uninstalled my IDE now that agents exist".
Why?
Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
Why?
Do you not understand that the IDE you deleted will slowly evolve into the industrialised agent harness you're now building yourself.
Maybe people aren't resigning from OpenAI and Anthropic because they foresee some mythical AGI doom scenario. Maybe they're resigning because they're finding the working environment miserable, chaotic and souless.
September 9, 2026 at 5:46 PM
Maybe people aren't resigning from OpenAI and Anthropic because they foresee some mythical AGI doom scenario. Maybe they're resigning because they're finding the working environment miserable, chaotic and souless.
Something new from the man who launched Vibe (not vibe coding - Vibe came first), with some help from his friends. Check it out.
I’ve been working on something new..
It’s called Statement
The internet is becoming increasingly difficult to trust.
Fake reviews. Paid influence. AI-generated content flooding our feeds.
As a new…...
I’ve been working on something new..
It’s called Statement
The internet is becoming increasingly difficult to trust.
Fake reviews. Paid influence. AI-generated content flooding our feeds.
As a new ...
www.linkedin.com
September 9, 2026 at 7:33 AM
Something new from the man who launched Vibe (not vibe coding - Vibe came first), with some help from his friends. Check it out.
Reposted by Jon Ayre
Sixty years ago today, the starship Enterprise launched and with it, a vision of the future where humanity had moved beyond its divisions. I was there at the beginning, and I can tell you...
September 8, 2026 at 5:30 PM
Sixty years ago today, the starship Enterprise launched and with it, a vision of the future where humanity had moved beyond its divisions. I was there at the beginning, and I can tell you...
Time to make a statement
Time to make a Statement | Jon Ayre
Time to make a Statement
lnkd.in
September 8, 2026 at 6:45 PM
Time to make a statement
Reposted by Jon Ayre
Dancing very very VERY close to saying “OpenAI stole our almost-complete proof of Navier-Stokes from internal access to our ChatGPT sessions because they were desperate to scoop Anthropic” which is absolutely 100% a believable thing OpenAI would do right now cims.nyu.edu/~tristanb/st...
September 8, 2026 at 7:13 AM
Dancing very very VERY close to saying “OpenAI stole our almost-complete proof of Navier-Stokes from internal access to our ChatGPT sessions because they were desperate to scoop Anthropic” which is absolutely 100% a believable thing OpenAI would do right now cims.nyu.edu/~tristanb/st...
Reposted by Jon Ayre
Presumably any system that does something other than showing you just the people you have chosen to follow in chronological order. Anything that does something different should be classed as publishing and the publisher should be fully legally liable for all content.
Australians will be able to switch off social media algorithms under planned legislation
www.bbc.co.uk/news/article...
www.bbc.co.uk/news/article...
Australians will be able to switch off social media algorithms under planned legislation
Firms like Meta, Google and TikTok should face "substantial" penalties if they don't give the option, minister says.
www.bbc.co.uk
September 7, 2026 at 7:36 AM
Presumably any system that does something other than showing you just the people you have chosen to follow in chronological order. Anything that does something different should be classed as publishing and the publisher should be fully legally liable for all content.
Is he delusional or disingenuous? Yet another example of Hinton declaring LLMs to be intelligent. No-one who has actually built a feedforward network or understands how one works can possibly think this, anymore than a person who built a tape recorder can believe there's an orchestra inside it.
Is he delusional or disingenuous? Yet another example of Hinton declaring LLMs to be intelligent. No-one who has actually built a feedforward network or understands how one works can possibly think… |...
Is he delusional or disingenuous? Yet another example of Hinton declaring LLMs to be intelligent. No-one who has actually built a feedforward network or understands how one works can possibly think th...
www.linkedin.com
September 7, 2026 at 6:55 AM
Is he delusional or disingenuous? Yet another example of Hinton declaring LLMs to be intelligent. No-one who has actually built a feedforward network or understands how one works can possibly think this, anymore than a person who built a tape recorder can believe there's an orchestra inside it.
As the internet tradition demands:
"Banana for scale"
(Love a good house spider - what should I name her?)
"Banana for scale"
(Love a good house spider - what should I name her?)
September 6, 2026 at 9:23 PM
As the internet tradition demands:
"Banana for scale"
(Love a good house spider - what should I name her?)
"Banana for scale"
(Love a good house spider - what should I name her?)
Reposted by Jon Ayre
Just moving billions of dollars in gold out of north America to London due to generalised ‘geopolitical unrest’. Carry on - nothing to see here
www.bbc.co.uk/news/article...
www.bbc.co.uk/news/article...
DNB: Netherlands bank moves billions in gold to London in 'crisis preparedness' move
The Dutch central bank removes some of its gold reserves from the US and Canada, saying it is because of "increasing geopolitical unrest".
www.bbc.co.uk
September 2, 2026 at 8:58 PM
Just moving billions of dollars in gold out of north America to London due to generalised ‘geopolitical unrest’. Carry on - nothing to see here
www.bbc.co.uk/news/article...
www.bbc.co.uk/news/article...
Reposted by Jon Ayre
i’m giving bsky a chance again. maybe it’s my own fault that i followed a bunch of folks that mainly only post about Serious Thinking Thoughts so im gonna unfollow everyone and start over, equally prioritizing serious things and unserious things.
August 31, 2026 at 7:17 PM
i’m giving bsky a chance again. maybe it’s my own fault that i followed a bunch of folks that mainly only post about Serious Thinking Thoughts so im gonna unfollow everyone and start over, equally prioritizing serious things and unserious things.
Reposted by Jon Ayre
Absolute masterclass in how NOT to write about AI.
ITS NOT HUMAN YOU CAN’T ATTRIBUTE HUMAN MOTIVATIONS AND EMOTIONS TO CODE.
ITS NOT HUMAN YOU CAN’T ATTRIBUTE HUMAN MOTIVATIONS AND EMOTIONS TO CODE.
August 29, 2026 at 10:48 AM
Absolute masterclass in how NOT to write about AI.
ITS NOT HUMAN YOU CAN’T ATTRIBUTE HUMAN MOTIVATIONS AND EMOTIONS TO CODE.
ITS NOT HUMAN YOU CAN’T ATTRIBUTE HUMAN MOTIVATIONS AND EMOTIONS TO CODE.
The irony of those who rightfully criticise the untrustworthy nature of AI writing, but then trust the same AI to "detect" AI writing. AI detectors are little more than magic eight balls, and if you judge a person or make important decisions based on their outputs, you're an irresponsible fool.
August 27, 2026 at 10:46 AM
The irony of those who rightfully criticise the untrustworthy nature of AI writing, but then trust the same AI to "detect" AI writing. AI detectors are little more than magic eight balls, and if you judge a person or make important decisions based on their outputs, you're an irresponsible fool.
Did a dog write this?
A single-celled parasite carried by cats, Toxoplasma gondii, infects an estimated third of all people on Earth, forms tiny cysts in the human brain, and in rodents famously rewires them to lose their ...
A parasite living in your brain may be subtly reshaping your personality, making you bolder, riskier, and less cautious than you'd otherwise be.
spacedaily.com
August 27, 2026 at 7:41 AM
Did a dog write this?
The reason workplaces aren't neuro-inclusive is because the people in them don't want to share their spaces with people who are different. The job of creating inclusive spaces is that of changing people, & people are notoriously difficult to change. Most of the apparent change is surface level only.
The reason workplaces aren't neuro-inclusive is because the people in those workplaces don't want to share their spaces with people who are different to them. The job of creating inclusive spa...
The reason workplaces aren't neuro-inclusive is because the people in those workplaces don't want to share their spaces with people who are different to them. The job of creating inclusive spaces of a...
www.linkedin.com
August 26, 2026 at 7:07 AM
The reason workplaces aren't neuro-inclusive is because the people in them don't want to share their spaces with people who are different. The job of creating inclusive spaces is that of changing people, & people are notoriously difficult to change. Most of the apparent change is surface level only.
There seems to be a common misperception about generative AI, neural networks and determinism. The "common wisdom" is that all neural networks are non-deterministic, including biological brains. This is fundamentally wrong.
There seems to be a common misperception about generative AI, neural networks and determinism. The "common wisdom" is that all neural networks are non-deterministic, including biological bra...
There seems to be a common misperception about generative AI, neural networks and determinism. The "common wisdom" is that all neural networks are non-deterministic, including biological brains. This ...
www.linkedin.com
August 23, 2026 at 9:23 AM
There seems to be a common misperception about generative AI, neural networks and determinism. The "common wisdom" is that all neural networks are non-deterministic, including biological brains. This is fundamentally wrong.