Zendoric

Zendoric

AI analysis and perspective, every day.

Updated: July 14, 2026
🎙️ Listen to today's podcast (in Spanish)

🔥 In focus today

When even the economists at OpenAI and Anthropic sign the alarm over jobs

🕒 Published on Zendoric: July 14, 2026 · 00:03

More than 200 economists, including 16 Nobel laureates and the chief economists of OpenAI and Anthropic, have signed an open letter comparing AI's disruption to the Industrial Revolution, but faster. They call for action now, before layoffs and an obsolete social safety net collide head-on.

The McConnell photo nobody believed: when not even a real image is enough against suspicion of AI

🕒 Published on Zendoric: July 14, 2026 · 00:03

After the death of Senator Lindsey Graham, rumors about Mitch McConnell's health forced his team to release a photo as proof of life. It didn't work: part of the public deemed it AI-generated and turned to chatbots to 'verify' it, with results that only fueled the doubt.

The Apple-OpenAI Trade-Secret Suit Is Serious. The Musk-Altman Name-Calling Isn't

🕒 Published on Zendoric: July 14, 2026 · 00:03

Apple has sued OpenAI over alleged theft of trade secrets, and within days the story collapsed into a schoolyard exchange of "scam" insults between Musk and Altman. Our take: separate the real legal question from the ego theater, because only one of them matters.

Anthropic's Qwen Distillation Claim Exposes the Real Weak Spot: API Defenses, Not China

🕒 Published on Zendoric: July 14, 2026 · 00:03

Anthropic alleges Alibaba's Qwen lab ran 28.8 million queries through ~25,000 fake accounts to distill Claude's coding skills. The accusation is unproven and geopolitically loaded — but the uncomfortable truth it reveals is that per-account rate limiting failed for six weeks straight.

📰 More AI analysis

The U.S. creates an 'AI coordinator' against childhood cancer: the cure isn't in the law, it's in the data

🕒 Published on Zendoric: July 14, 2026 · 00:03

A bipartisan bill in the U.S. Congress seeks to appoint a federal AI coordinator to accelerate pediatric cancer research and force the interoperability of clinical data. The intent is laudable and the political consensus is real, but the fine print —$100 million a year and no new regulatory power— says a lot about just how hard this problem really is.

Claude Code no longer reads the web, it sees it: the built-in browser and the race to control the coding agent

🕒 Published on Zendoric: July 14, 2026 · 00:03

Anthropic embeds a Chromium browser inside Claude Code so the agent can interact with live websites, not just static screenshots. It's a modest technical move that reveals something bigger: the battle to control the 'plumbing' of agentic AI in software development.

Four freshmen build in 10 weeks what a factory couldn't afford for years

🕒 Published on Zendoric: July 14, 2026 · 00:03

A university project in Oakland replaced whiteboards and paper at a machining plant in Santa Clara with an app featuring automatic translation into five languages. The most striking figure —an estimated $180,000 in annual savings— matters less than who was able to build it, and how fast.

'AI washing' now has case law: the SEC doesn't need new laws to go after AI hype

🕒 Published on Zendoric: July 14, 2026 · 00:03

Two recent U.S. rulings confirm that exaggerating AI capabilities to investors fits squarely within old-fashioned securities fraud, with no need for specific rules. The case links Monolithic Power Systems and Omnicare to growing legal scrutiny of AI marketing.

9 out of 10 HR leaders regret AI layoffs: the bill for cutting by title, not by talent

🕒 Published on Zendoric: July 14, 2026 · 00:03

A survey of 600 HR leaders reveals that nearly 90% would handle their AI-related layoffs differently: a third lost critical skills and more than a third have already rehired for half of the eliminated roles. The problem isn't that AI is useless, but how it was decided who to cut.

Nate: why I keep opening GPT-5.6 Sol even though Fable 5 is the smartest model

🕒 Published on Zendoric: July 14, 2026 · 00:03

Nate, who benchmarks AI models for a living, confesses something he finds uncomfortable in that role: last night he opened ChatGPT 5.6 Sol even though he considers Fable 5 the smartest model.

Apple sues OpenAI over alleged theft of trade secrets for its AI hardware

🕒 Published on Zendoric: July 14, 2026 · 00:03

Apple has filed a lawsuit against OpenAI in a federal court in Northern California, accusing the artificial intelligence lab of trade secret theft.

A Nobel-winning chemist leaves UC Berkeley to lead an AI institute at Tsinghua, China

🕒 Published on Zendoric: July 14, 2026 · 00:03

Omar Yaghi, winner of the 2025 Nobel Prize in Chemistry, is leaving his post at the University of California, Berkeley to lead a new artificial intelligence institute at Tsinghua University in Beijing, as the Chinese university itself announced.

Meta pulls its controversial Instagram AI feature that let users generate images from other people's public accounts

🕒 Published on Zendoric: July 14, 2026 · 00:03

Meta has backtracked on an artificial intelligence feature in Instagram that had sparked a strong backlash among users and talent agencies. The tool, launched this very week as part of a package of new AI features, is no longer available, the company itself has confirmed.

False positive: this 'AGI' isn't superintelligence, it's a Canadian silo maker

🕒 Published on Zendoric: July 14, 2026 · 00:03

The AI news tracker picked up a corporate earnings release that has nothing to do with artificial general intelligence: here 'AGI' is the ticker for Ag Growth International, an agricultural equipment company. We flag it for the sake of editorial honesty, not as a relevant finding.

Three votes and a unanimous rejection: what a tiny local poll reveals about AI in the classroom

🕒 Published on Zendoric: July 14, 2026 · 00:03

An opinion poll by a Canadian local newspaper on whether elementary schoolchildren should use generative AI in school received only three votes, and all three were a resounding 'no.' The sample is minuscule, but the reflection it offers —families' instinctive wariness toward AI for children— is real and worth reading carefully.

Canada regulates agentic AI in banking: when the supervisor puts a name to the risk (Claude Mythos included)

🕒 Published on Zendoric: July 14, 2026 · 00:03

OSFI, Canada's financial regulator, has published six practices to curb the risks of generative and agentic AI at banks and insurers. And according to an email revealed by Reuters, it went so far as to explicitly cite Anthropic's Claude Mythos when warning about cyber risks: the scrutiny now has a model's name.

The brain decides earlier than we thought: the clue that explains why today's AI burns so much energy

🕒 Published on Zendoric: July 14, 2026 · 00:03

A PNAS study on mice and their whiskers finds that the primary sensory cortex takes part in the decision, not just receives it, thanks to feedback signals from frontal regions. The feedforward architecture that today's neural networks copied —CNNs and transformers included— may be leaving on the table the efficiency the brain achieves on 20 watts.

53.6% of Australian university assignments used AI: the answer isn't to chase it, it's to redesign what gets assessed

🕒 Published on Zendoric: July 14, 2026 · 00:03

Turnitin detected AI in more than half of university submissions in Australia between October 2025 and April 2026, and Anthropic ranks the country as the world leader in per capita use of Claude. The figure is alarming, but the relevant question isn't how much AI is in a text, it's what learning is left behind it.

VERAXA taps AI to better pick cancer targets: the small, real step versus the big headline

🕒 Published on Zendoric: July 14, 2026 · 00:03

VERAXA Biotech (Nasdaq: VRXA) is leaning on Ardigen's bioinformatics to prioritize target combinations in its BiTAC(R) bispecific antibody platform, aiming to reduce the toxicity that sinks many oncology drugs in the clinic. It's a limited and reasonable use of AI in biotech, though the wrapper —a microcap investor-relations release— calls for keeping your feet on the ground.

When the corporate message falls out of alignment, AI acts as a coherence auditor (not a spokesperson)

🕒 Published on Zendoric: July 14, 2026 · 00:03

Two Tuck (Dartmouth) professors propose using AI to detect when a company's messaging to investors, employees and customers stops adding up to a coherent story, using BlackRock's discursive shift on ESG as a case study. Their thesis: AI is useful for diagnosing, not for deciding the message.

The Illinois teacher accused of using AI to create child abuse exposes the real limit of safeguards

🕒 Published on Zendoric: July 14, 2026 · 00:03

An Illinois teacher, accused of secretly recording students aged 12 to 14 and using AI to create sexually explicit images, could be fired this week. The case, still at the indictment stage, exposes how easily AI today generates child abuse material from real photos.

Circle's CEO imagines an economy where AI agents sign contracts and pay on their own with stablecoins

🕒 Published on Zendoric: July 14, 2026 · 00:03

Jeremy Allaire, co-founder of Circle, publishes an essay in which AI makes thinking and working cheaper, and blockchain makes transacting cheaper: agents with their own wallet would negotiate, sign and pay in stablecoins without going through a human department. It's a thesis, not a roadmap — and the person signing it also sells the money that would underpin it.

Illinois bans AI from judging teachers: the first legal barrier against teacher-evaluation software

🕒 Published on Zendoric: July 14, 2026 · 00:03

Governor JB Pritzker signed a law barring school administrators from using AI to evaluate teachers, and teachers from using it to meet their performance goals. It's one of the first legal red lines on which decisions an algorithm cannot make in school.

Cloudflare and AWS put a price on the web: agentic AI starts paying for every click it makes

🕒 Published on Zendoric: July 14, 2026 · 00:03

Cloudflare is opening a gateway for websites to charge AI crawlers and agents for each access, and AWS adds its WAF with Coinbase. It's the first serious attempt to turn automated traffic into a microtransaction market, with a deadline already set: September 15.

The 'easy gold' of AI-made kids' cartoons is a mirage: COPPA and YouTube already charge up to an 80% toll

🕒 Published on Zendoric: July 14, 2026 · 00:03

'Side hustle' tutorials promise easy money generating children's cartoons with AI in minutes. What they leave out: COPPA cuts advertising by up to 80%, YouTube has already shut down channels with 4.7 billion views, and pediatricians document real harm to child development.

FOD#158: If we must act now on AI, what should we really do?

🕒 Published on Zendoric: July 14, 2026 · 00:03

The editorial in this edition of Turing Post starts from recent news: more than 200 economists, AI researchers and tech leaders —among them at least 16 Nobel laureates and figures from OpenAI, Anthropic and Google— have signed a declaration titled "We Must Act Now."

How Miro is redefining AI collaboration in product development

🕒 Published on Zendoric: July 14, 2026 · 00:03

This email from McKinsey & Company, signed by Basel Kayyali and Florian Niedermann, global leaders of McKinsey Technology, promotes a piece from its "New from McKinsey & Company" series focused on Miro, the collaborative digital whiteboard platform.

Anthropic discovers a "J-space" of internal thoughts in Claude

🕒 Published on Zendoric: July 14, 2026 · 00:03

This edition of The Algorithm (MIT Technology Review) revolves around new mechanistic interpretability research published by Anthropic, the company behind Claude and, according to the email, the world's most valuable AI company, with a valuation nearing one trillion dollars…

Distillation in the era of LLMs: when the student started to answer

🕒 Published on Zendoric: July 14, 2026 · 00:03

This TheSequence Knowledge article opens by revisiting the famous 2015 distillation paper (Hinton et al. on "dark knowledge"), but not to talk about its temperature trick or that notion of dark knowledge, rather to point to something deeper: the conceptual world that paper took for granted.

Apple sues OpenAI over alleged theft of trade secrets for its hardware project

🕒 Published on Zendoric: July 14, 2026 · 00:03

Apple filed a lawsuit against OpenAI on Friday, accusing the artificial intelligence company of stealing trade secrets as part of an effort to build its own hardware device.

OpenAI and Google, linked to sanctioned affiliates: the original article could not be accessed

🕒 Published on Zendoric: July 14, 2026 · 00:03

It was not possible to access the actual content of this article. The page downloaded from the Yahoo Finance link only returned the cookie notice and the site's privacy options, without any journalistic text on the announced topic.

Cursor develops 'Sand', a general AI agent to take on Anthropic's Claude Cowork

🕒 Published on Zendoric: July 14, 2026 · 00:03

Transparency notice: the original The Information article on this topic is largely blocked behind a paywall. The downloaded content only includes the headline, the site navigation and an unrelated context line ("Before Apple Lawsuit, a Growing Unease at iPhone Maker Over OpenAI"), without the body…

Senator Ed Markey introduces a package of bills to regulate data centers, automated hiring and chatbots

🕒 Published on Zendoric: July 14, 2026 · 00:03

Massachusetts Democratic Senator Ed Markey has introduced what he calls an "AI accountability agenda," a set of about a dozen bills aimed at curbing the harmful effects of artificial intelligence in the United States.

Grok 4.5 leads coding benchmarks (content unavailable)

🕒 Published on Zendoric: July 14, 2026 · 00:03

It was not possible to access the actual content of this YouTube Short. What was downloaded from the link is only YouTube's standard cookie consent and sign-in screen (privacy notices, language selector, "Accept all" / "Reject all" buttons), without any audio fragment…

SK Hynix raises $26.5 billion in the largest foreign IPO in U.S. history

🕒 Published on Zendoric: July 14, 2026 · 00:03

SK Hynix, the South Korean memory chip maker, has pulled off the largest stock market debut by a foreign company in Wall Street history. The company announced it raised $26.5 billion (40 trillion South Korean won) in its U.S. IPO, surpassing the $25 billion…