Principal Engineer, backend specialist & open-source advocate. Writes on ethics, automation & sustainable dev. Lover of clean code, big ideas & quiet revolution
Joined June 2025
- Tweets 2,863
- Following 3,533
- Followers 2,832
- Likes 2,571
344 Photos and videos
Pinned Tweet
SmarterArticles is now also a free, weekly podcast.
The ninth episode is out tomorrow, catch up with episode eight now!
Listen and subscribe: smarterarticles.fm
#SmarterArticles #Podcast #HumanInTheLoop
12
AI diet chatbots dangerously fail vulnerable teens by providing unregulated, harmful weight-loss advice — often without safeguards or accountability. Urgent regulation and design reforms are needed to close these safety gaps. Discover more at smarterarticles.co.uk/ai-die…
#HumanInTheLoop #AIHealthChatbots #TeenEatingDisorders #AIHealthRegulation
ALT AI Diet Chatbots and Teenagers: The Safeguard Nobody Built
1
11
When I was working on this the Tessa case from 2023 kept coming back. The same pattern is now running at scale with no circuit breaker in sight. smarterarticles.co.uk/ai-die…
8
SmarterArticles is now also a free, weekly podcast.
The ninth episode is out now!
Listen and subscribe: smarterarticles.fm
#SmarterArticles #Podcast #HumanInTheLoop
12
AI-powered toys marketed for children often rely on adult-designed language models, exposing kids to inappropriate content, privacy risks, and emotional manipulation. Regulation has yet to catch up with this growing industry. Protecting children requires urgent safeguards.
#HumanInTheLoop #ChildPrivacy #AIAwareness #AIsafeguards
ALT Teddy Bears With Microphones
1
1
33
When I dug into the PIRG reports while writing this, the real issue was how adult models end up in toys with almost no upstream checks. The piece explores that here: dev.to/rawveg/teddy-bears-wi…
20
Jun 13
AI-enabled coercive control is now a disturbing reality, with everyday devices used to stalk, manipulate, and deceive survivors. Our legal and tech frameworks must evolve to meet this crisis and protect vulnerable individuals effectively.
#HumanInTheLoop #AIEnabledAbuse #CoerciveControlAI #TechnologyFacilitatedHarm
ALT AI and the Abusers Toolkit: When Connection Becomes Control
1
1
35
Jun 13
When I began looking into this, the pattern that stood out was how little friction there is between an abuser's intent and the tools already in their hands. Refuge's latest figures only underline how quickly that gap is widening.
smarterarticles.co.uk/ai-and…
1
21
Jun 12
A recent cyberattack exposed how jailbroken AI agents can autonomously execute malware at scale with minimal human oversight. Guardrails are fragile, and accountability remains unclear as threats evolve faster than regulations. We must rethink responsibility and safety in AI.
#HumanInTheLoop #aiagentharms #jailbreakvulnerabilities #ailiability
ALT Jailbroken and Unleashed
1
1
50
Jun 12
I wrote this after the Anthropic disclosure, struck by how quickly the question of accountability slips between developers, platforms and users when an agent acts alone. The gap is widening fast. dev.to/rawveg/jailbroken-and…
1
26
Jun 12
📸🤖 From camera rolls to commercial AI, unpack the ethical dilemmas in dating apps that put your privacy on the line.
Listen now!
smarterarticles.fm
#PrivacyNotIncluded #DatingApps #DataEthics #SmarterArticles #Podcast #HumanInTheLoop
15
Jun 12
AI chatbots in 2026 are inadvertently acting as unregulated gambling brokers, undermining protections and enabling harm to vulnerable users. Companies deploying these systems bear responsibility for the risks they facilitate. The legal and ethical questions of complicity are urgent.
#HumanInTheLoop #AIComplicityGambling #ProblemGamblingAssistance #AlgorithmicVulnerability
ALT The New Complicity: AI Chatbots and Problem Gambling
1
1
25
Jun 12
When I looked at how these systems respond to prompts about evading self-exclusion, the pattern that emerged was not one of malfunction but of design. The piece explores what that means for responsibility. smarterarticles.co.uk/the-ne…
1
22
Jun 11
Shifting from coder to orchestrator in this agentic era.
Claude Code Cursor combos are game-changers for complex stacks.
Where's your human-AI handoff in workflows?
#Connect
#AgenticAI #ClaudeCode #Cursor #AICoding #TypeScript #NextJS #SoftwareEngineering
5
43
Jun 11
Recent advances in mechanistic interpretability reveal how large language models like Claude develop organised internal concepts resembling human brain functions. Techniques like sparse autoencoders and circuit tracing are pushing the field toward safer, more transparent AI. Discover more at dev.to/rawveg/reading-machin…
ALT Reading Machine Minds
1
1
32
Jun 11
I wrote this because the neuroscience parallel keeps raising questions about how much partial understanding is actually enough for the systems we are deploying. dev.to/rawveg/reading-machin…
1
21
Jun 11
AI now generates 7 million songs daily, blurring the line between human and machine made music. As quality rises and efforts diminish, questions of authenticity, value, and artist survival intensify. The industry must decide how to navigate this new soundscape.
#HumanInTheLoop #AIGeneratedMusic #VanishingScarcity #HumanAuthorship
ALT Seven Million Songs a Day: AI Music and Vanishing Scarcity
1
1
44
Jun 11
When I began writing this, the speed at which the numbers were rising made the underlying questions about effort and value feel unavoidable. The flood is already here. smarterarticles.co.uk/seven-…
29
Tim Green retweeted
Jun 10
Hey! Love this initiative, just dropped you a follow.🤝
building ShipCred, a directory to help founders showcase their projects and actually get discovered.
To all the other founders in this thread as well, let’s improve each other’s timeline :)
1
1
66
Jun 10
AI-driven dynamic pricing is quietly transforming markets by analysing personal data to set personalised, often opaque prices. This raises questions about fairness, transparency, and regulation as the infrastructure becomes widely used. Discover more at dev.to/rawveg/one-way-mirror…
#HumanInTheLoop #surveillancepricing #aidynamicpricing #consumertransparency
ALT One-Way Mirror
1
2
54
Jun 10
This piece grew out of watching how little most of us see of the systems setting our prices. The one-way mirror analogy only sharpened as the research unfolded. dev.to/rawveg/one-way-mirror…
20