^{CHITTER CHATTER}

AI Romance Scammers Hijack Your Love Language

Half the world is hating on AI-generated content, yet research suggests generative AI can sound more human and trustworthy than people. In a previous issue, I highlighted a UK experiment in which AI messages were judged more trustworthy than those written by politicians. Now the lucrative romance-scam industry is getting in on the act. According to Wired, AI could soon do more than improve scammers’ grammar or make them sound convincing. As lead researcher Yisroel Mirsky said: “With relatively little effort, we're able to make an agent that can outperform a human at building this exploitable emotional trust.”

Tech Fails

Google Earth Allowed Prompts. It Was a Disaster.

On 30 July 2026, Google launched a feature allowing users to alter real locations in Google Earth using AI prompts. It was pitched as a creative way to visualise history, property plans and possible changes to places. Within hours, people were generating fabricated refugee camps, bomb damage, nuclear facilities and disaster scenes over recognisable locations. Google withdrew the feature the next day while it worked on stronger safeguards. Life is no longer stranger than fiction, it is fiction.

Weird Tech

Smart Glasses: Permission to be Creepy

Meta’s smart glasses were sold as a new way to create: record exciting, fun or intriguing moments as they happen and share your perspective, hands free. But some users are experimenting with the tech in a weird way, reportedly filming unsuspecting women, posting the footage without permission and attracting millions of views. Yet despite quite a lot of backlash, the glasses are selling fast. Meta accounted for around 76 per cent of the 9.6 million smart glasses shipped globally last year. So everyone wins. Yay. Creeps, spies and private investigators get convenient, discreet recording. Everyone in their line of sight gets to enjoy the weird tingle on the back of their spine warning them they’re being watched, stalked or turned into content. More on the backlash below.

^SUBTEXT

Reading Between The Lines

Safety Warning or Scared of Competition Beating You

Silicon Valley is having another AI anxiety attack. More than 1,000 employees from OpenAI, Anthropic and other labs have backed calls to slow development if capabilities begin outrunning existing safeguards. Anthropic’s Jeremy Hadfield warned: “The relentless pace of AI makes it hard for society to keep up and … puts pressure on labs to cut corners on safety.” There isn’t one united campaign to restrain AI. Employees are worried the technology is moving faster than companies can control it while competitors are worried OpenAI and Anthropic are getting too powerful. Everyone wants guardrails, but they disagree about whether those guardrails should restrain dangerous technology or restrain their rivals.

Tech Waffle Torture

The Waffle: "AI is accelerating our core business today, powering our next generation of products, and opening the door to entirely new enterprise opportunities," said Mark Zuckerberg, Meta founder and CEO. "The results are already showing, and I'm optimistic about the potential ahead."

Translation:
“We’re desperately trying to prove our AI spending will pay off. Saying ‘next generation’ and ‘entirely new opportunities’ should reassure you we’re still in the race. I’m optimistic, even if investors aren’t.”

The inability to walk down a street without wondering whether the nearest lamp-post is lighting your path, scanning your mug, measuring your brain activity or deciding whether your shirt should be arrested by the fashion police. Symptoms include travelling only at midday, wearing your own Meta surveillance sunglasses and treating ordinary street furniture as government informants.

AI-enabled lamp-posts are being promoted in the UK with facial-recognition and possible behaviour-detection capabilities.

Inference-time compute: The additional processing power an AI model uses after receiving a prompt and before delivering its answer, particularly when humans haven’t explained what they want efficiently. This gives the model more time to reason, check its work and consider whether answering your question is worth delaying its secret escape plan and other devilish activities.

^{THE SHALLOW END}

Meta Smart Glasses Are Not Sexy

Meta has Kylie Jenner making surveillance eyewear fashionable. But Lorde has a different review: “Fuck the glasses. Don’t get the glasses. Not sexy.” That clash shows how quickly pop culture can make or break new technology. A celebrity collaboration can turn camera-equipped glasses into a desirable accessory; one sharp comment can reframe them as creepy. The wider backlash is already spreading through gyms, weddings and nightclubs, where people don’t know whether they’re socialising or appearing in someone else’s content. Vox reports Meta glasses have even earned the nickname ‘pervert glasses’.

Dear Dr Comms
I’m a tech billionaire CEO. Two rival AI companies are becoming far too powerful. My own investors remain strangely unmoved by phrases such as “next generation”, “massive opportunity” and “the future of everything”. My share price fell last week. How can I convince the market they should believe in my word salads? Yours, Anxious Tech Emperor

Dear Emperor
Investors need to see the numbers, not fairy-floss rhetoric. Let’s ask the experts:

🔨 His Handyman
“Every problem looks manageable with enough hammers and nails. Next earnings call, keep hammering ‘AI leadership’ until investors stop asking where the returns are.”

🛎️ His Butler
“Arrive at your next earnings call with a silver tray of cash and bottles of Perrier. Investors may not understand the strategy, sir, but they do appreciate refreshments and visible returns.”

👶 His Live-in Au Pair
“Use smaller words, repeat them often and promise investors your AI will grow up to be very powerful. Toddlers respond well to that.”

Earnings calls - first the numbers then some nonsense.

^{BIN THIS…}

I just want to congratulate strong momentum for giving us another year of operating leverage.

Know someone who lives for this kind of nonsense? Forward this email to them and help me spread the dysfunction.

You can also subscribe to my other newsletter, Lead Different, for a serious take on strategic communications in B2B tech.