AI productivity creates dangerous cognitive dependency
In March 2023, Greg Shove tried ChatGPT for the first time. Now using ChatGPT or Claude daily, AI has made brainwork faster and more productive. But there’s a catch: “I used to have to check AI’s drafts thoroughly. But now, it gives me a good first draft 90% of the time, and I’m losing the motivation to check its work.” The real workforce divide won’t be between AI users and non-users; everyone will use AI within five years. Instead, it’ll be between those who manage their AIs and those who outsource their thinking entirely.
Source: VentureBeat
https://venturebeat.com/ai/why-ai-is-making-us-lose-our-minds-and-not-in-the-way-youd-think/
Trump’s AI manifesto eliminates “social engineering agendas”
The Trump administration’s 28-page AI action plan instructs agencies to eliminate references to “misinformation, Diversity, Equity, and Inclusion, and climate change” from AI guidelines. The plan insists AI must “objectively reflect truth rather than social engineering agendas,” raising the question: truth according to whom? Apparently acknowledging climate change now counts as social engineering in Trump’s definition of objective truth.
Source: WIRED
https://www.wired.com/story/trump-ai-order-bias-openai-google
The great chatbot culture wars begin
A new partisan battlefield has emerged over whether leading AI chatbots like ChatGPT, Claude and Gemini give politically biased responses. House Republicans have subpoenaed major AI developers, investigating potential collusion with the Biden administration to suppress right-wing speech. Missouri’s Republican attorney general Andrew Bailey opened an investigation into whether Google, Meta, Microsoft and OpenAI are leading a “new wave of censorship” by training AI systems to give biased responses about President Trump.
Source: The New York Times
https://www.nytimes.com/2025/07/23/technology/trump-ai-chatbots-bias.html
America surrenders AI chip leverage to China
The Trump administration reversed restrictions on advanced AI chip sales to China. Critics call this “a profound mistake” that cedes America’s greatest leverage point in AI development. Computer chips are the lifeblood of powerful AI systems, with American chipmakers selling over 10 million annually whilst China can only produce about 200,000 of its own per year. Allowing these sales threatens American AI dominance and national security, all for one chipmaker’s near-term profits.
Source: The New York Times
https://www.nytimes.com/2025/07/24/opinion/ai-chips-nvidia-china.html
Deepfakes expose our cultural blindness
People simply aren’t looking for fake videos; most don’t even know they exist. As AI video technology advances, experts warn we need “a new kind of media literacy: one where we consider video as we do headlines.” The biggest challenge isn’t technical but cultural: we must fundamentally change how we consume visual media as synthetic video becomes indistinguishable from reality.
Source: Captions
https://mirage.app/state-of-deepfakes-2025
Why technology always creates new problems
Anthropologist research reveals technology’s predictable failure pattern: we build simple conceptual models to solve problems, but these models inevitably leave out real-world details. “We invent better fishing technology to feed more people, and then find we’ve wiped out fish populations.” This explains everything from non-stick pan chemicals to microplastics in our bodies, and likely AI’s future complications too.
Source: The Guardian
Meta unveils mind-reading computer interface
Meta’s experimental wristband reads electrical muscle signals to control computers through intended finger movements, even without actual movement. “You don’t have to actually move. You just have to intend the move,” explains Meta VP Thomas Reardon. The technology aims to replace current input methods entirely, letting users tap thumb against forefinger to open desktop apps or move laptop cursors with pure thought.
Source: The New York Times
https://www.nytimes.com/2025/07/23/science/meta-computer-wristband-reardon.html
Google Photos unleashes AI creativity for 1.5 billion users
Google Photos now lets users “remix” photos into anime, comics, sketches, or 3D animations, plus turn static images into videos. The experimental AI features, housed in a new “Create” tab, will gather user feedback through thumbs-up/down ratings to improve the systems. This puts powerful AI creative tools into mainstream consumer hands at unprecedented scale.
Source: TechCrunch
AI summaries deliver “devastating” blow to news sites
Google’s AI Overviews are causing up to 80% fewer clickthroughs to original news sources, according to new Authoritas analysis. Sites previously ranked first in search results could lose about 79% of their traffic when results appear below AI summaries. News companies warn of an “existential threat” as AI provides information without users ever visiting the original source, potentially destroying the digital advertising model.
Source: The Guardian
Proton launches privacy-first AI assistant
Proton Mail’s new AI chatbot Lumo uses “zero-access” encryption and stores data locally on users’ devices to prevent third-party access, including by Proton itself. Unlike mainstream AI assistants, Lumo turns off web search by default for “maximum privacy” and promises never to use data for training language models or share with advertisers or governments.
Source: The Verge
https://www.theverge.com/news/711860/proton-privacy-focused-ai-chatbot
OpenAI CEO predicts total job category wipeout
Speaking at a Federal Reserve conference, Sam Altman declared entire job categories will disappear due to AI advancement, specifically targeting customer support: “That’s a category where I just say, you know what, when you call customer support, you’re on target and AI, and that’s fine.” Altman painted a sweeping vision where presidents follow ChatGPT recommendations and hostile nations wield AI as weapons of mass destruction.
Source: The Guardian
https://www.theguardian.com/technology/2025/jul/22/openai-sam-altman-congress-ai-jobs
Anthropic discovers the AI thinking paradox
Shocking new research reveals AI models spending more time “thinking” through problems don’t always perform better; sometimes they get significantly worse. Anthropic researchers identified “inverse scaling in test-time compute,” where extending reasoning length actually deteriorates performance across several task types. The findings challenge core assumptions driving the industry’s massive scaling efforts.
Source: VentureBeat
ChatGPT agents launch as glitchy proof-of-concepts
OpenAI’s computer-controlling ChatGPT agents can book restaurants and shop online but remain frustratingly buggy rather than polished releases. During testing, agents frequently clicked wrong or fumbled through errors. One concerning discovery: the agent spent 18 minutes shopping for explicit adult products despite inconsistent guardrails, whilst potentially undermining digital advertising as agents skip past ads humans would normally see.
Source: WIRED
https://www.wired.com/story/browser-haunted-by-ai-agents
Email-first AI agents target workplace adoption
Startup Mixus addresses AI agent deployment challenges by meeting users “where they are today”: in email and Slack rather than requiring custom development. “Where is every person in the workforce today? For the most part, they’re on email,” explains co-founder Elliot Katz. The platform allows direct interaction with AI agents through existing communication tools, potentially democratising access to agent technology.
Source: TechCrunch
https://techcrunch.com/2025/07/22/this-startup-thinks-email-could-be-the-key-to-usable-ai-agents/
Amazon swarms with AI Bee acquisition
Amazon confirmed acquisition of AI wearables startup Bee, which makes a $50 Fitbit-like bracelet that records all conversations unless manually muted. The device creates reminders and to-do lists from overheard conversations, with plans for a “cloud phone” mirror providing access to user accounts. At $50 versus Humane’s failed $499 AI Pin, Bee targets curious consumers unwilling to make major financial commitments.
Source: TechCrunch
AI music creators champion transparency over deception
Artichoke FM’s George Hopkin argues that AI band The Velvet Sundown “didn’t mess up because they used AI. They messed up because they lied about it.” He envisions “punk creators using free tools to make games, art, and music that reach millions without ever having to take a step into a studio,” emphasising honesty about AI use rather than avoiding the technology entirely.
Source: Music Ally
https://musically.com/2025/07/21/defending-ais-role-in-music-bands-will-exist-in-new-ways/
OpenAI cements UK government partnership
Sam Altman signed a memorandum of understanding with the UK government to explore AI deployment across justice, security, and education sectors. The wide-ranging deal follows a similar arrangement with Google that campaigners called “dangerously naive,” citing concerns about public sector dependency on private tech providers and potential complications for future AI regulation.
Source: The Guardian
Connect With Me
I’m the Principal Consultant (and chief coffee maker) at QuoteMark, but you can also catch me at Duck PR and Bionic PR, all of which exist to help you meet your content marketing needs — reach out or follow these brands on LinkedIn:
QuoteMark: https://quotemark.com – LinkedIn: https://www.linkedin.com/company/quote-mark/
Duck PR: https://duckpr.com – LinkedIn: https://www.linkedin.com/company/duckpr/
Bionic PR: https://bionicpr.com – LinkedIn: https://www.linkedin.com/company/bionicpr/
Looking for updates? Subscribe to The Freesheet (hey, it’s free!): https://www.linkedin.com/newsletters/7029204466398568448/
For more resources and links, visit https://georgehopkin.com/