- MediaMorph
- Posts
- MediaMorph Edition 66 - by HANA News
MediaMorph Edition 66 - by HANA News
Some summer cynicism and a psychosis warning
Mark Riley
August 26, 2025
In partnership with
MediaMorph Edition 66 - by HANA News
Some summer cynicism and a psychosis warning
Was this newsletter forwarded to you? Sign up here
The written-by-a-human bit
We are zooming out this week to discuss MIT’s bombshell report that apparently 95% of generative AI pilots are failing, then we will zoom in to a couple (of ours) that aren’t.
MIT’s claim that 95% of generative pilots fail was so inflamatory that certain AI related stocks fell in an already febrile environment. On closer inspection, however, it appears the methodology was seriously flawed, as exposed by Nathaniel Whittlemore in The AI Daily Brief:
In short, the true value is being seen at the individual level with widespread adoption of tools like ChatGPT, often without oversight. Where strategic projects do fail, it is not the underlying technology to blame, but resistance to change in the organisation. Execs may be misaligned with staff, data might not be ready, or funding is pulled before value is realised. After 15 years of driving innovation in large companies, I see nothing new here. Change management has always been hard. The difference today is that AI will find its way into the building whether it is invited or not. Pulling pilots on the back of this spurious report would be cowardly and counter-productive.
Meanwhile, in the plus ledger, we are doing our best to crack the local news business-model-problem - we put this Bristol newsletter together every Friday in about an hour (Edition 48) with a sister publication in Oxford (https://oxford.hana.news/) and plans for 10 more cities — we use our own homegrown AI software https://app.hana.news/. The audience is growing 10% month on month organically.
What we have learned:
💡 Human-in-the-loop - everything is human-checked and human-edited
💡 Honour the source - attribute and promote the story, drive traffic back
💡 Use humans to lead with a human editorial - report the pulse
💡 Build a community of civic folk who want to improve their city
💡 Use AI to find the stories in the source data - courts, council, universities, police logs, health data
There are plenty of AI-cowboys out there pumping out slop on the backs of hard-won journalism - we are looking to find the balance between efficiency and efficacy.
On a micro-level, we aim to demonstrate that responsible AI use can empower local journalism, serving as a tool rather than a sledgehammer. My point is that when used responsibly, AI can transform the economics of journalism, improve the product and leave us humans to do the engaging, meaningful, civic stuff.
Read about another successful case study from the editor of Cleveland.com and The Plain Dealer - Our 2025 journalism has sharpened, deepened and grown, thanks to AI: Letter from the Editor
Finally, for those still on the beach and with time to ponder the bigger existential questions, Mustafa Suleyman, CEO of Microsoft AI, and the co-founder and former head of applied AI at DeepMind, has raised the alarm about people seeing consciousness in AI where none exists, leading to “psychosis risk”. There is a problem no one asked for. I am always grateful to the billionaire architects of AI when they sound the alarm about their own creations - a theme I explore in my book The AI Prophet for release in paperback on September 1st.
Mark Riley, CEO Mathison AI
PS on topic - check out the inaugural episode of The FixEd podcast, where host Eliza Diamantidi speaks with Jakub Parusinski about AI's impact on media
Hi At Mathison AI, we are firm advocates for media companies building their own subscriber-facing chatbots for deeper engagement. The FT, Aftonbladet, Forbes, The Washington Post and Rappler are all leading the way. We understand the need for RAG-based training, guardrails and guidelines. We build intelligent AI that has the inbuilt humility to admit when it doesn’t know the answer. Speak to us today about testing your archive in a safe environment. Mark Riley [email protected] |
AI and Journalism
I Tested How Well AI Tools Work for Journalism Cjr - AI tools are revolutionising journalism by enhancing reporting efficiency through automated content generation, data analysis, and audience engagement. However, their integration raises ethical concerns about authenticity, bias, and misinformation, necessitating a careful balance between innovation and journalistic integrity. |
Tests Show That Top AI Models Are Making Disastrous Errors When Used for Journalism Futurism - August 23, 2025 Media executives' increasing reliance on AI for content creation is leading to a surge in poorly executed outputs that harm credibility and drive readers away, according to a recent investigation by NYU’s Hilke Schellmann. As concerns grow about the reliability of AI-generated journalism, the industry faces an urgent need to balance innovation with the preservation of trust and quality. |
New Jersey bill seeking to restrict media use of AI draws sharp criticism • The Jersey Vindicator The Jersey Vindicator - August 21, 2025 New Jersey lawmakers have introduced a bill mandating news organizations to label AI-generated content and credit original sources, imposing hefty fines for violations, while media scholar Jeff Jarvis argues the proposal could hinder journalistic innovation and reflects an unnecessary panic about technology's role in reporting. The Jersey Vindicator emphasizes the importance of community support for its journalism efforts, highlighting how donations directly contribute to its sustainability and quality coverage. |
Our 2025 journalism has sharpened, deepened and grown, thanks to AI: Letter from the Editor Cleveland - August 23, 2025 Local journalism is experiencing a renaissance thanks to advancements in artificial intelligence, which enhance content quality and expand coverage. Tools like "Editors Eyes" are empowering reporters to produce clearer, more accurate stories while enabling deeper exploration of topics, proving that AI is a valuable ally rather than a replacement for journalists. |
Wired, Business Insider delete phony articles allegedly written by AI under alias ‘Margaux Blanchard’ New York Post - August 22, 2025 Freelance journalist Margaux Blanchard has had multiple articles retracted by major outlets like Wired and Business Insider after allegations surfaced that her work was AI-generated and contained fabrications, including a fictional mining town and unverifiable sources. The situation highlights growing concerns over journalistic integrity in an age of rapidly advancing technology. |
AI's role in Ukrainian wartime media Deutsche Welle - August 21, 2025 In the midst of escalating conflict in Sumy, Ukraine, independent media outlets like Cukr are harnessing AI technologies to combat disinformation while navigating severe staffing shortages and mental strain among journalists. Tools like Osavul are crucial for real-time analysis of false narratives, enabling media to focus on delivering accurate stories that restore hope amidst the chaos. |
Why media must try to hold on to its credibility in the age of artificial intelligence South China Morning Post - August 19, 2025 AI-generated content has surged in quality, particularly in mainland China, with platforms like Baidu showcasing its impact on social and current affairs. This evolution is not only fostering discussions about "AI intimacy" but also raising concerns over job market disruptions as society braces for transformative changes driven by AI technology. |
FixEd podcast: Jakub Parusinski on AI in the media industry The Fix - August 21, 2025 In the inaugural episode of The FixEd podcast, host Eliza Diamantidi speaks with Jakub Parusinski about AI's impact on media, emphasising that while AI can produce grammatically correct content, it lacks the depth and quality necessary for true journalism. Parusinski warns that as essential reporting functions are overtaken by other platforms, young journalists must adapt to a changing landscape where traditional training tasks are being diminished. |
AI Is Taking Over Our Social Media Feeds, but Maybe Not How You Expect CNET - August 22, 2025 As generative AI transforms social media management, with 96% of social media managers relying on these tools for tasks like content creation and customer service, the challenge lies in balancing efficiency with authenticity to avoid "AI slop." Professionals emphasize the need for human creativity and connection to maintain trust and engagement amidst growing concerns over content quality and inauthenticity. |
Find out why 1M+ professionals read Superhuman AI daily.
In 2 years you will be working for AI
Or an AI will be working for you
Here's how you can future-proof yourself:
Join the Superhuman AI newsletter – read by 1M+ people at top companies
Master AI tools, tutorials, and news in just 3 minutes a day
Become 10X more productive using AI
Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.
AI and Academic Publishing
3 in 5 first-years report weekly use of generative AI to complete assignments The Duke Chronicle - August 21, 2025 The Class of 2028 at Duke University shows a notable increase in generative AI usage for academic work, with 60% using it weekly—up from 49.3% among the previous class—primarily for problem sets and coding assignments. The survey highlights distinct patterns in AI usage across disciplines, revealing that engineering and natural sciences students utilize AI more frequently than those in social sciences and arts. |
Omnivore Presents: SciDish | August 2025: Research Integrity in Scholarly Publishing Ift - Intense pressure to "publish or perish" in academia is pushing some researchers towards unethical practices, compromising the integrity of research and the credibility of institutions. A growing movement advocates for systemic changes that prioritize the quality and impact of research over mere publication quantity, aiming to cultivate a more ethical academic culture. |
How reliable are AI book summaries? A closer look Qazinform.com - August 24, 2025 AI-generated summaries can quickly offer insights into nonfiction works but often overlook the depth of fiction, risking oversimplification and loss of nuance. As AI's role in literature expands, concerns arise about diminished diversity in style and interpretation, potentially stifling creativity and originality in future works. |
AI Assignment Agents TrendHunter.com - August 19, 2025 Grammarly has unveiled eight innovative AI agents, including the AI Grader, designed to enhance student writing by offering personalized feedback and predictive grading. This trend highlights a significant shift towards AI-powered academic tools that improve writing processes and support educational outcomes for the upcoming school year. |
How Wiley's New AI Alliances Could Shape WLY's Future in Academic and Technology Research Simplywall - August 23, 2025 Under CEO Matt Kissner, John Wiley & Sons is leveraging AI partnerships with firms like Anthropic and Amazon Web Services to enhance its Research and Learning segments, projecting $1.8 billion in revenue by 2025. While the company's innovative approach promises growth, the inherent volatility of AI licensing revenues introduces risks to earnings visibility. |
‘Business-as-usual’ offers from publishers raise walk-away fears Times Higher Education (THE) - August 20, 2025 UK universities may sever ties with major journal publishers due to proposed agreements featuring unacceptable price hikes, prompting institutions to seek reductions amid tightening budgets. As they prepare for potential disruptions, many are enhancing inter-library loan services and exploring alternative arrangements while negotiations continue. |
Editorial: AI and Inverse Methods for Building Digital Twins in Neuroscience Recent advancements in deep learning are revolutionising neuroscience and healthcare, enhancing clinical diagnostics and aiding neurosurgeons with applications such as seizure prediction and brain connectivity mapping. Prominent studies showcase the effectiveness of neural networks in analysing medical imaging, decoding brain activity, and improving our understanding of complex neurological processes. |
This newsletter was partly curated and summarised by AI agents, who can make mistakes. Check all important information. For any issues or inaccuracies, please notify us here
View our AI Ethics Policy
