Mark Stockley
@ai-fix-mark.bsky.social
Host of the award-winning AI Fix podcast (@theaifix.show) and cybersecurity evangelist for ThreatDown.
⚠️ Never use an LLM to generate a password. ⚠️ A password is only safe if it’s hard to guess. That means no predictable patterns, no 123456, no p@ssword!, no dictionary words, no pet names, no favourite sports teams. It needs to be random. Something like: G7$kL9#mQ2&xP4!w
When it comes to AI you can have safe or honest, not both. Research suggests that the neurons stopping hallucinations are the same neurons stopping the AI from telling you how to make ricin. Want more honesty? It's easier to jailbreak. Want better guardrails? Sorry, it's a big fat liar now.
In episode 87 of The AI Fix, @ai-fix-mark.bsky.social and guest host Nik Roberts take yet another gulp from AI's infinite well of creepy things, and discover a robot that learned to lip sync in the mirror. Sadly, we can't confirm whether it did so while singing into a hairbrush.
Do you wish your AI was a bit less... demonic? A little less ghostly and a bit more consultant-y? Strangely, even if you don't, you probably do.
In episode 69 of @theaifix.show, Florida decides to solve its python problem with some laughably bad robot rabbits.
Has AI created the most annoying thing in the world? In episode 69 of The AI Fix, @grahamcluley.com and I discover Brain Rot, the Gen Alpha AI craze that's even more annoying than Crazy Frog ringtones.
In episode 64 of @theaifix.show, a robot folds some laundry, a robot crab looks for love on the beaches of Portugal, the “Godfather of AI” thinks our only hope is to build motherly AI, the UK government has a terrible idea, and our hosts discover a long-lost sixties soul hit, “Dusty Plasma”.
I'm asking GPT-5 to create some illustrations for me, and I'm starting to wonder if its much heralded "tool use" includes MS Paint.
Apparently, GPT-5's facility with "sustaining unrhymed iambic pentameter" is going to make it better at helping you with "everyday tasks like drafting and editing reports." As long as those reports don't contain charts, I guess.
Large Language Models (LLMs) don't work the way we thought. LLMs are sometimes described as next-word predictors or likened to a fancy autocomplete. Recent research from Anthropic shows that they are far more than that.
In our next episode, we find out how AIs often vomit up unpleasant stuff. Follow "The AI Fix" in all podcast apps. 🟢 Spotify: theaifix.show/spotify 🟣 Apple Podcasts: podcasts.apple.com/gb/podcast/t... 🛑 Pocket Casts: pca.st/itunes/17533... Everywhere else: theaifix.show/follow/ #podcast #ai
Who is Mary Lee Pfeiffer's son—and what has that got to do with AI?
In the next episode of "The AI Fix" podcast... Is your AI suffering from anxiety???
Elon Musk's AI is more left wing than Mark Zuckerberg's, and way, way closer to AOC and Bernie Sanders than it is to Trump.
Want a visual clue as to what we will be talking about in episode 39 of "The AI Fix" podcast?
AIs really, really don't want to draw six-fingered gloves. An @theaifix.show listener asked two AIs to "create an image of a leather glove from the time of the three musketeers novel. This glove needs to have six fingers. It should be lying on a lace doily on a table." Count the fingers.
Only 10% of people have a higher IQ than ChatGPT. Four of OpenAI's models now have an above average IQ, according to the Tracking AI website, and one of them is above 120. Which begs the question, what if Artificial General Intelligence (AGI) has already been invented, and we just didn't notice?
🦠🦠🦠 "The DeepSeek model did the worst of any model we've ever tested [on national security evaluations, like providing info on bioweapons]" 🦠🦠🦠 In this week's episode of @theaifix.show, I look at the appalling state of DeepSeek security.
I saw a post saying KIMI is the new DeepSeek. Time to roll out the rigorous @theaifix.show benchmark.
In episode 36 of The AI Fix, @grahamcluley.com and I take a long look at DeepSeek R1, the upstart AI taking on OpenAI's o1.
How it feels to publish an AI podcast today that was edited on Sunday 😂
Ho ho ho... Guess who has a special Christmas episode of their #AI podcast for you... Dropping down your chimmney at 3pm UK, in all good podcast apps.
For this week's special Christmas episode of @theaifix.show, I asked ChatGPT to plot an optimal delivery route for Santa. After years of being ignored, the kids of Eouth Asia finally get a look in!
ChatGPT and Sora down worldwide. Clearly someone needs an AI Fix… ahem. hackread.com/sora-chatgpt...
Sora and ChatGPT Currently Down Worldwide: OpenAI Working on a Fix
Follow us on Bluesky, Twitter (X) and Facebook at @Hackread
hackread.com
Aww shucks! We love it when we get a nice review. And we reckon this one wasn't written by a bot! If you enjoy The #AI Fix, and want to make our lives feel like they have some purpose as we prepare ourselves for The Singularity™, do leave us a note on Apple Podcasts, Spotify, Podchaser etc ❤️
In our next episode of "The AI Fix" podcast, we'll be taking a trip into the world of #JamesBond Out on Tuesday 3 December, at 3pm UK! 🟢 Spotify: theaifix.show/spotify 🟣 Apple Podcasts: podcasts.apple.com/gb/podcast/t... 🛑 Pocket Casts: pca.st/itunes/17533...
