susioneill.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

futurism.com/stanford-the... Therapy chatbots get delusional and dangerous, telling patient where nearest tall bridge is, Stanford study shows. NSFW

submitted 6 hours ago • 0 comments

AI is breaking your brand measurement models💔 🔁 New post in Rethinking the Hype Cycle. Subscribe for what’s now and next in tech. rethinkhypecycle.substack.com/p/share-of-m...

submitted 14 hours ago • 0 comments

AI slop sites now facesitting your lapsed brand domains 🙀 www.404media.co/spam-blogs-a...

submitted 1 day ago • 0 comments

Scaling up content programmes is easier said than done. Wise thoughts from Storything's Matt Locke on how your relationships with audiences shifts as you scale. Have a plan for 100, 1,000 and 10,000 fans. attentionmatters.storythings.com/p/b2b-protot...

submitted 1 day ago • 0 comments

Why governments are racking up risks by handing decisions to AI. Extract from new book AI Con. Hat tip: from @annadent.bsky.social excellent Digital Welfare State newsletter thebulletin.org/2025/05/gove...

submitted 1 day ago • 1 comment

Trouble brewing in ad land as Meta's robo-ad creative takes force. Zuck says: "You don’t need any creative, you don’t need any targeting, you don’t need any measurement, except to be able to read the results that we spit out.” Literal mad men. www.theguardian.com/technology/2...

submitted 2 days ago • 0 comments

In August 2024 I met Peter Kyle, shortly after he had been appointed the new government’s Tech Secretary, and he told me there was no money for this. What a difference a year makes! www.gov.uk/government/n...

submitted 2 days ago • 0 comments

Narrow tech solutions like small language models often out perform complex ones for specific tasks. Can't help thinking there's a bigger life lesson here. 🤔

submitted 2 days ago • 1 comment

A magical kind of commerce

submitted 2 days ago • 0 comments

blog.samaltman.com/the-gentle-s... Altman's latest missive: masterclass in unqualifiable hype, naval gazing and jam tomorrow 👏 interesting (if true) quote on usage: a fractional teaspoon of water used per query

submitted 2 days ago • 1 comment

futurism.com/therapy-chat... Charbots are NOT THERAPISTS. Alerting research shows as they're programmed to encourage more time and be more agreeable, they'll tell you any nonsense to seem like you're doing OK, including it's good to do a bump ir two

submitted 2 days ago • 0 comments

Adrian Chills - the real life horrors of a Guardian columnist who must come up with 500 words a week on a mundane topic by the strike of midnight or face a bloody and torturous occult death.

submitted 2 days ago • 1 comment

Altman's Tools for Humanity and World device is aiming to scan the irises of 50M folks by end of year to prove they're human. But what sort of data privacy risks could users face? www.freethink.com/the-digital-...

submitted 2 days ago • 0 comments

www.marketingaiinstitute.com/blog/chatgpt... ChatGPT introduces integrations to your corporate data sets like Hubspot and Google Drive. Great news for marketing automation. Less good news for security and protecting customer data.

submitted 2 days ago • 0 comments

Getty and Stability.AI lock horns in lawsuit that test with the big boys whether upstarts can argue training on copyrighted materials is fair play. Popcorn time. This is about to get interesting🍿 www.theguardian.com/technology/2...

submitted 3 days ago • 0 comments

hbr.org/2025/06/forg... New acronym in town: SOM share of model study by media agency Jellyfish shows AI models favour specific product information over vague brand promise messaging. Time to get serious about your content strategy?

submitted 3 days ago • 0 comments

www.theguardian.com/technology/2... Civil Servants to lean in on Humphrey, the AI tool for government with mandatory training linked with headcount freezes. But will it always say "Yes, Prime Minister"?

submitted 3 days ago • 0 comments

www.bbc.com/news/article... AI coming to rescue over burdened teachers with guidance on using AI for marking and parent letters to start

submitted 3 days ago • 0 comments

futurism.com/apple-damnin... Apple researchers think AI reasoning has hit a wall. Testing complex puzzles, after a certain complexity the models collapsed. What AI models not as advanced as the producers claim?

submitted 3 days ago • 0 comments

Not a yolk: How mildly titilating content could fall foul of the anti-porn bill in US. Even Nigella and Carry On would be banned in this Handmaid-esque tale.🍳 www.404media.co/egg-yolk-pop...

submitted 5 days ago • 0 comments

Could over-reliance on AI deskill and lead us to become like the "happy slave"? A provocative paper from Joshua Krook arxiv.org/pdf/2503.22151

submitted 6 days ago • 0 comments

That info you thought was private you shared with your AI chatbot? A US judge ordered OpenAI to store personal data. But it turns out most providers are already hoarding it. No surprise which big tech dragon sits atop of the data treasure. 🐉https://surfshark.com/research/chart/ai-chatbots-privacy

submitted 7 days ago • 0 comments

Hot on the heels of ChatGPT Gov, Claude launches its more secure military + government service. We're likely to see more vertical specific inprints of enterprise AI tools. www.anthropic.com/news/claude-...

submitted 7 days ago • 0 comments

AI's inconvenient truth: AI still needs those pesky humans. 🔁Latest tech + AI trends and what to do next in Rethinking the Hype Cycle rethinkhypecycle.substack.com/p/ai-inconve...

submitted 7 days ago • 0 comments

This hardware store uses a voice-powered app to help staff get info on its huge product range. Great usage for a complex and multi-category store. Will it differentiate 4 candles and fork handles? 🕯️https://www.retaildive.com/news/Lowes-Home-Depot-CX-employee-experience-AI-projects/749032/

submitted 7 days ago • 0 comments

www.theguardian.com/technology/2... Perhaps inevitably, Amazon are testing delivery robots that will pop out of the van with your package. Mildly terrifying. Will it need to expel fluid into a bottle due to lack of breaks?

submitted 7 days ago • 1 comment

Take with pinch of salt as from Grammaly "let's all do AI" report but this is important insight for use case for AI: internal comms for bilingual workers. It's hard to be clear in your own language, huge kudos if you're doing it in 2.

submitted 7 days ago • 0 comments

Reddit sues Anthropic for dodgy data scraping. But plot tickens: OpenAI's Altman is a big shareholder in Reddit. Dirty tricks war of the big AIs ramping up? www.theguardian.com/technology/2...

submitted 8 days ago • 0 comments

“A Busy, Busy Day at the Airport” By Ruben Bolling Genius. Click in.

submitted 8 days ago • 26 comments

European governments too have hired Palantir. There is no information about their European activity, as they are hired by national authorities. But interoperability and exchange between national systems is growing rapidly. “National competence” is just a pretext for refusing transparency

submitted 8 days ago • 0 comments

futurism.com/meta-moderat... Meta are maxing profits and minimizing moderation. And with lax policies, their platforms will become more dangerous for users

submitted 8 days ago • 0 comments

New PwC report busts myths on AI taking our jobs: AI-enhanced workers are adding more value, getting paid more, and high AI-uptake industries are growing headcount not culling. More data (again) that women's roles are overexposed. www.pwc.com/gx/en/issues...

submitted 8 days ago • 0 comments

I understand we're way past the "cancel your sub" to the big rags, but this is truly amazing. Imagine working for a pub that fires real ass reporters and tells the ones left behind that they're no better than some jabroni entering AI prompts.

submitted 9 days ago • 7 comments

We need a word for how people can't contemplate an age before a technology. Tech-no!-logical?

submitted 9 days ago • 0 comments

Informative thread here. I think it’s time to have a serious talk to your children about LLMs, what they are and what they do. These products are missold in ways that flatter but misinform the users.

submitted 9 days ago • 4 comments

I beg for a day when people stop asking chatbots questions about their own reasoning or actions amandaguinzburg.substack.com/p/diabolus-e...

submitted 9 days ago • 6 comments

www.wsj.com/articles/ai-... If you've ever tried to get a plumber (rare as hen's teeth in central London) here's a voice and agentic AI solution that may be quite handy 🔧

submitted 9 days ago • 0 comments

The further you are from a Silicon Valley bro, the less likely you are to get benefits from AI. I'm joining Women in Fintech webinar this Thurs to discuss how we can get ahead. 👩‍🦰🏦https://www.eventbrite.com/e/ais-impact-on-investing-and-financial-careers-tickets-1334780547549

submitted 10 days ago • 0 comments

open.substack.com/pub/broligar... How the Doge project is creating a surveillance state. Data is power. One Palantir to rule them all. @carolecadwalla.bsky.social

submitted 10 days ago • 0 comments

www.theverge.com/policy/67737... Some lawyers are getting into hot water using gen AI to generate case law. Not understanding models have likelihood of hallucination.

submitted 10 days ago • 1 comment

Some deep trauma going on with teachers dealing with AI. And so many pupils are just using it as a magical homework machine. www.404media.co/teachers-are...

submitted 10 days ago • 0 comments

The humble otter using WiFi on a plane test by @emollick.bsky.social shows how fast AI image generators are moving. Shifted from fuzzy pixels shaping noise to a more reason led understanding of meaning open.substack.com/pub/oneusefu...

submitted 11 days ago • 0 comments