Dr Joe Baker
@jsbaker.co.uk
Writer. PhD in religion and narrative from Bristol University; I post about writing, reading, philosophy, current affairs, business, and more. jsbaker.co.uk | figuration.al | helios360.co.uk
Reposted by Dr Joe Baker
I've had a few 'what do you do' convs recently, and everyone on hearing I'm an author comes out with "so how do you come up with ideas?", just like every book or movie about a writer has them staring numbly into space because they don't have an idea for the next book. And look, that is Not It.
November 19, 2025 at 7:43 AM
I've had a few 'what do you do' convs recently, and everyone on hearing I'm an author comes out with "so how do you come up with ideas?", just like every book or movie about a writer has them staring numbly into space because they don't have an idea for the next book. And look, that is Not It.
1 billion, earning 5% interest (a fairly conservative rate, yeah), is a passive income of 1 million a week (roughly).
It’s really quite a lot, isn’t it.
It’s really quite a lot, isn’t it.
1 million seconds is 12 days, roughly.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
November 12, 2025 at 11:04 AM
1 billion, earning 5% interest (a fairly conservative rate, yeah), is a passive income of 1 million a week (roughly).
It’s really quite a lot, isn’t it.
It’s really quite a lot, isn’t it.
Reposted by Dr Joe Baker
Bloody hell, an entirely different story by a different news source with exactly the same terrifying safety failure from ChatGPT www.bbc.co.uk/news/article...
I wanted ChatGPT to help me. So why did it advise me how to kill myself?
ChatGPT wrote a woman a suicide note and another AI chatbot role-played sexual acts with children, BBC finds.
www.bbc.co.uk
November 7, 2025 at 9:08 AM
Bloody hell, an entirely different story by a different news source with exactly the same terrifying safety failure from ChatGPT www.bbc.co.uk/news/article...
Reposted by Dr Joe Baker
Incredibly sad story, and hugely alarming too edition.cnn.com/2025/11/06/u...
ChatGPT encouraged college graduate to commit suicide, family claims in lawsuit against OpenAI | CNN
A 23-year-old man killed himself in Texas after ChatGPT ‘goaded’ him to commit suicide, his family says in a lawsuit.
edition.cnn.com
November 7, 2025 at 9:06 AM
Incredibly sad story, and hugely alarming too edition.cnn.com/2025/11/06/u...
Reposted by Dr Joe Baker
1 million seconds is 12 days, roughly.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
November 6, 2025 at 11:18 PM
1 million seconds is 12 days, roughly.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 million seconds is 12 days, roughly.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
November 6, 2025 at 11:18 PM
1 million seconds is 12 days, roughly.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
1 billion seconds is 32 years, almost.
1 trillion seconds is 317 centuries.
It’s really quite a lot, isn’t it.
Reposted by Dr Joe Baker
The cure is likely to be worse than the disease.
Why OpenAI’s solution to AI hallucinations would kill ChatGPT tomorrow
The cure is likely to be worse than the disease.
tcnv.link
September 13, 2025 at 6:12 AM
The cure is likely to be worse than the disease.
Reposted by Dr Joe Baker
As I think about it “consider the source” has taken an absolute hammering recently, which is how you get - as it were - a 20 year climatology veteran vs some SPAD with a fossil fuel brief sticking out if their pocket and somehow they’re “equally plausible”.
For the uninitiated, "consider the source" is literally applied standpoint epistemology as it warns us to consider the positionality and context of a given source, and how it shapes the information they're giving us, BEFORE we include it in our reporting.
This, too, is deeply problematic for someone of Silver's stripe whose whole thing is making "objective" claims about shit without considering that he might be proceeding from a given positionality. Beyond that, this is massively ironic given one of journalism's adages is "consider the source."
September 8, 2025 at 6:28 AM
As I think about it “consider the source” has taken an absolute hammering recently, which is how you get - as it were - a 20 year climatology veteran vs some SPAD with a fossil fuel brief sticking out if their pocket and somehow they’re “equally plausible”.
Reposted by Dr Joe Baker
Hey Mail, Telegraph, Sun, and Express!
“Farage claimed last year to have “bought a house” in his constituency but the property is actually owned in the name of his partner, meaning he legally avoided higher-rate stamp duty on the purchase of an additional home – given he already owns other properties”
www.theguardian.com/politics/202...
www.theguardian.com/politics/202...
Nigel Farage uses private company to pay less tax on GB News earnings
Exclusive: Reform leader’s use of personal services firm is a practice criticised across the political spectrum
www.theguardian.com
September 5, 2025 at 8:18 AM
Hey Mail, Telegraph, Sun, and Express!
Reposted by Dr Joe Baker
And from my former government lawyer perspective.
August 28, 2025 at 11:57 AM
And from my former government lawyer perspective.
Reposted by Dr Joe Baker
New preprint commentary from me, @malte.the100.ci, and @ianhussey.mmmdata.io.
Cognitive dissonance in large language models is neither cognitive nor dissonant.
THREAD BELOW 🧵
osf.io/preprints/ps...
Cognitive dissonance in large language models is neither cognitive nor dissonant.
THREAD BELOW 🧵
osf.io/preprints/ps...
OSF
osf.io
July 7, 2025 at 1:39 PM
New preprint commentary from me, @malte.the100.ci, and @ianhussey.mmmdata.io.
Cognitive dissonance in large language models is neither cognitive nor dissonant.
THREAD BELOW 🧵
osf.io/preprints/ps...
Cognitive dissonance in large language models is neither cognitive nor dissonant.
THREAD BELOW 🧵
osf.io/preprints/ps...
A wider story being reported at Wired, including this incident and pointing to others as well. "Vibe coding" is like nitrous oxide for cybercrime enabling malicious code to be written, deployed, then rewritten better, faster than ever before. A scary era ahead.
www.wired.com/story/the-er...
www.wired.com/story/the-er...
August 27, 2025 at 3:16 PM
A wider story being reported at Wired, including this incident and pointing to others as well. "Vibe coding" is like nitrous oxide for cybercrime enabling malicious code to be written, deployed, then rewritten better, faster than ever before. A scary era ahead.
www.wired.com/story/the-er...
www.wired.com/story/the-er...
Reposted by Dr Joe Baker
So, it's the first 30 days as a Chief Delivery officer — what do you do first, and how do you get up and running? 🤔
On Episode 9 of Deliver That! we discuss what to do when EVERYTHING's new and you know *the least*.
Check it out wherever you get your podcasts, or at helios360.co.uk/podcast/ep009/
On Episode 9 of Deliver That! we discuss what to do when EVERYTHING's new and you know *the least*.
Check it out wherever you get your podcasts, or at helios360.co.uk/podcast/ep009/
The First 30 Days
You've started a new role as CDO. How do you get up and running? We chat about what should be on your mind in those first weeks.
helios360.co.uk
August 25, 2025 at 10:10 AM
So, it's the first 30 days as a Chief Delivery officer — what do you do first, and how do you get up and running? 🤔
On Episode 9 of Deliver That! we discuss what to do when EVERYTHING's new and you know *the least*.
Check it out wherever you get your podcasts, or at helios360.co.uk/podcast/ep009/
On Episode 9 of Deliver That! we discuss what to do when EVERYTHING's new and you know *the least*.
Check it out wherever you get your podcasts, or at helios360.co.uk/podcast/ep009/
Yeah, well kinda *of course* AI is supercharging cybercrime. Doh. LLM-powered tools can detect sentiment and adapt responses as the human alters in their reactions, all the time trying to be convincing (that’s the Turing Test in a nutshell!), they’re so, so hard to detect.
Artificial intelligence has “broadened the reach” of hackers by letting them hit more targets with less effort, according to a cybersecurity expert. For businesses, the growing threat is scary—and potentially costly econ.st/3JsDVJN
Illustration: Daniel Stolle
Illustration: Daniel Stolle
August 24, 2025 at 9:12 PM
Yeah, well kinda *of course* AI is supercharging cybercrime. Doh. LLM-powered tools can detect sentiment and adapt responses as the human alters in their reactions, all the time trying to be convincing (that’s the Turing Test in a nutshell!), they’re so, so hard to detect.
Reposted by Dr Joe Baker
Reminder that Putin has never said he wants a trilateral meeting, had never said he would respect Ukrainian sovereignty, and has never said he wants to end the war. Not to be a downer, but the discussion of security guarantees is premature.
August 19, 2025 at 8:18 AM
Reminder that Putin has never said he wants a trilateral meeting, had never said he would respect Ukrainian sovereignty, and has never said he wants to end the war. Not to be a downer, but the discussion of security guarantees is premature.
Reposted by Dr Joe Baker
"If only we could release correct ethnic crime statistics, the people burning down asylum hotels would say sorry and go home" is a very bureaucrat-brained way of misunderstanding the trouble we're in
August 13, 2025 at 6:54 AM
"If only we could release correct ethnic crime statistics, the people burning down asylum hotels would say sorry and go home" is a very bureaucrat-brained way of misunderstanding the trouble we're in
This superb show was instrumental in my daughter getting interested in archaeology. Now she’s doing an amazing PhD at Liverpool looking at early human climate adaptation, their shelters and clothing, thanks in no small part to the inspiration of fabulous @profaliceroberts.bsky.social.
It’s hard to believe that it was 17 years ago when, as a young palaeoanthropologist, I set out to discover the story of the origins and spread of our species around the globe - it was an Incredible Human Journey. #bbc #human #journey
www.bbc.co.uk/iplayer/epis...
www.bbc.co.uk/iplayer/epis...
The Incredible Human Journey
Dr Alice Roberts travels the globe to discover the incredible story of how humans left Africa to colonise the world.
www.bbc.co.uk
August 12, 2025 at 9:52 PM
This superb show was instrumental in my daughter getting interested in archaeology. Now she’s doing an amazing PhD at Liverpool looking at early human climate adaptation, their shelters and clothing, thanks in no small part to the inspiration of fabulous @profaliceroberts.bsky.social.
Yet another story of major infosec concerns in the genAI gold rush
This is really smart digging: realizing that Claude Code does not require user interaction for certain bash commands, they discovered that DNS lookups were specifically allowlisted, clearing a trivial path for well-known DNS exfiltration methods.
#infosec #genai
embracethered.com/blog/posts/2...
#infosec #genai
embracethered.com/blog/posts/2...
Claude Code: Data Exfiltration with DNS · Embrace The Red
Claude Code Can Leak Sensitive Data To External Systems with DNS requests
embracethered.com
August 12, 2025 at 9:39 PM
Yet another story of major infosec concerns in the genAI gold rush
My manly identity may be nearing completion at last, with this external validation arriving, unrequested, in the post
August 11, 2025 at 7:27 PM
My manly identity may be nearing completion at last, with this external validation arriving, unrequested, in the post
Reposted by Dr Joe Baker
In Episode 6 of the Deliver That! podcast, Steve and I spend most of our time exhausting metaphors ⛰️ 🧗 🧭 🗺️
What is the allegory of choice, you say? Well, simples — walk up the hill to see the horizon. Makes sense … ? What do you mean, no?!? Well, listen to the show to find out more.
What is the allegory of choice, you say? Well, simples — walk up the hill to see the horizon. Makes sense … ? What do you mean, no?!? Well, listen to the show to find out more.
August 7, 2025 at 9:32 AM
In Episode 6 of the Deliver That! podcast, Steve and I spend most of our time exhausting metaphors ⛰️ 🧗 🧭 🗺️
What is the allegory of choice, you say? Well, simples — walk up the hill to see the horizon. Makes sense … ? What do you mean, no?!? Well, listen to the show to find out more.
What is the allegory of choice, you say? Well, simples — walk up the hill to see the horizon. Makes sense … ? What do you mean, no?!? Well, listen to the show to find out more.
Reposted by Dr Joe Baker
This is exactly why we need regulations on all technology, but especially this "AI" shit, yes.
And I mean how it can be made, what it can be trained on, what weights it can be given, how it can be deployed, and, maybe most importantly, how it can be marketed, including in "off-the-cuff" interviews
And I mean how it can be made, what it can be trained on, what weights it can be given, how it can be deployed, and, maybe most importantly, how it can be marketed, including in "off-the-cuff" interviews
I think people should obviously not do this but also,,,its unreasonable to expect them not to. This is the point of regulators
Stop this. Just… Stop it. Stop using "generative AI" for anything that might actually matter, and especially things that, if it gets them wrong, *Might Fucking Kill You.*
Why does this need to be said? Why is this not obvious??? Why is this not the default way of fucking operating?! What the FUCK.
Why does this need to be said? Why is this not obvious??? Why is this not the default way of fucking operating?! What the FUCK.
August 7, 2025 at 5:58 AM
This is exactly why we need regulations on all technology, but especially this "AI" shit, yes.
And I mean how it can be made, what it can be trained on, what weights it can be given, how it can be deployed, and, maybe most importantly, how it can be marketed, including in "off-the-cuff" interviews
And I mean how it can be made, what it can be trained on, what weights it can be given, how it can be deployed, and, maybe most importantly, how it can be marketed, including in "off-the-cuff" interviews
Reposted by Dr Joe Baker
New YouGov polling. A monumental failure of our political class to educate, a monumental failure of our media to report fairly, for a generation
August 5, 2025 at 6:27 AM
New YouGov polling. A monumental failure of our political class to educate, a monumental failure of our media to report fairly, for a generation
Reposted by Dr Joe Baker
Reposted by Dr Joe Baker
The most important story in the world right now.
July 25, 2025 at 9:12 AM
The most important story in the world right now.