AI analysis and perspective, every day.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Microsoft is instructing its sales force to compare OpenAI's and Anthropic's models unfavorably against its own, while replacing both with its in-house MAI AI in Excel and Outlook. The irony is obvious: it's attacking those who until recently were its strategic suppliers.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Microsoft is reportedly coaching its salesforce to talk down OpenAI, Anthropic and Google while pushing its in-house MAI models across Office. The bigger story isn't a sales script—it's the platform quietly swallowing the partners it once depended on.
🕒 Published on Zendoric: July 17, 2026 · 00:24
DARPA's DICE program envisions 100,000 AI agents negotiating tasks without a central brain, trading a single point of failure for swarm resilience. It's ambitious R&D, not deployed capability — and the real frontier here is governing autonomy, not counting agents.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Anthropic brings enterprise-grade agent capabilities —planning, browsing, coding— to a price far below its top-tier models. The effect isn't a smarter model, but a cheaper frontier: autonomous AI is no longer only for those who can afford it.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Hong Kong's new education blueprint puts AI at the center of school policy, but a recent survey reveals that its actual adoption varies enormously by subject. The document is a good starting point, but on its own it doesn't close that gap.
🕒 Published on Zendoric: July 17, 2026 · 00:24
A recent high school graduate in Pennsylvania, self-taught in Python and machine learning, has joined a Penn State pilot project using AI to improve the diagnosis of head and neck cancers. The story is small, but it points to something bigger: who can get into AI-driven medical research today.
🕒 Published on Zendoric: July 17, 2026 · 00:24
An NHS study with more than 1,600 prostate biopsies shows that AI, integrated into the real workflow, changed the diagnosis in 1 out of every 20 cases and cut waiting time by up to 30 hours. It doesn't replace the pathologist: it makes them faster and more accurate.
🕒 Published on Zendoric: July 17, 2026 · 00:24
State Street, one of the world's largest custodian banks, is aiming for $1 billion in benefits by combining AI with internal reorganization. The firm makes no secret that there will be staff cuts in the process.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Infobae published a short piece with Alan Daitch under the title 'Can AI become smarter than us?'. The available material does not include verifiable statements or data from the video, so at Zendoric we prefer to be honest about that limitation rather than fill in with generalities.
🕒 Published on Zendoric: July 17, 2026 · 00:24
At a colloquium at the Cuajimalpa unit, UAM academics discussed what remains of human authorship when generative AI intervenes in scripts, texts and teaching materials. The debate is not against technology, but about how to prevent it from diluting the responsibility of those who research, teach or create.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Klarna and other companies that boasted of replacing staff with AI are now paying to get them back. The pattern repeats: cut fast, overpromise, underestimate the human element, and spend years rebuilding what was thrown away.
🕒 Published on Zendoric: July 17, 2026 · 00:24
RT-One canceled the contract for a 96-hectare plot for its AI data center plan in Brazil after it emerged that the fund that owns it is under investigation for financial fraud linked to Banco Master. The case exposes a little-discussed risk of the AI infrastructure boom: due diligence on who controls the land and the energy.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Anthropic's CEO has personally donated one million dollars to Public First, the super PAC backing candidates who favor more AI safety guardrails. The figure comes amid a clash with Leading the Future, the rival PAC that already prevailed in a New York primary. The bill for who regulates AI is being paid, literally, in cash.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Match Group CFO Steven Bailey explains how the company behind Tinder and Hinge is becoming 'AI-native': nearly two-thirds of Tinder's improvements this year are algorithmic, and it has had to set token limits and slow hiring to afford it. Dating, increasingly mediated by software, now also has a bill to balance.
🕒 Published on Zendoric: July 17, 2026 · 00:24
The University of Colorado Boulder delayed student access to ChatGPT Edu in spring over privacy and cost concerns; now, before reactivating it in August, it will require completing a mandatory training module. It's the first visible sign that AI literacy is becoming an access requirement, not an option.
🕒 Published on Zendoric: July 17, 2026 · 00:24
A Kansas school district is adding artificial intelligence to its cameras to detect firearms in seconds and alert police. It's a real, measurable improvement in reaction time, but also a symptom: the technology arrives before the solution to the problem that made it necessary.
🕒 Published on Zendoric: July 17, 2026 · 00:24
It's no longer enough to protect the campaign's official inbox: attackers clone the candidate's voice to request urgent transfers and bombard staff with MFA notifications at 3 in the morning until an exhausted volunteer hits 'approve'. Election security is entering a phase where the weak link is no longer the server, but the exhausted human being.
🕒 Published on Zendoric: July 17, 2026 · 00:24
A new technical framework, CAVA, proposes translating the tangle of logs from different agentic AI environments into a single, verifiable format. It's a small but telling building block: the race is no longer just for more capable agents, but for being able to prove what they did and with what permission.
🕒 Published on Zendoric: July 17, 2026 · 00:24
This McKinsey Quarterly dispatch features a piece titled "Cost versus value: Managing agentic AI system performance," focused on how companies should measure what they actually pay for generative AI and, in particular, for agentic AI systems.
🕒 Published on Zendoric: July 17, 2026 · 00:24
The email announces a new McKinsey Global Institute report titled "Agents, robots, and us: How AI reshapes work and skills in Latin America," authored by Alexis Krivkovich, Anu Madgavkar, Jaime Szigethi, Fernando Sinagra, María Jesús Ramírez and Francesca Bencini.
🕒 Published on Zendoric: July 17, 2026 · 00:24
OpenAI has developed a language model specialized in hacking, dubbed GPT-Red, whose function is to serve as a sparring partner to strengthen the defenses of its other models against cyberattacks.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Elon Musk has quietly acquired APR Energy, a Jacksonville-based company that operates a fleet of trailer-mounted gas turbines and diesel/gas engines, with a combined capacity of over 1 GW.
🕒 Published on Zendoric: July 17, 2026 · 00:24
A hacker who breached Suno's systems, one of the largest AI music generation tools, shared source code and internal documents with the outlet 404 Media detailing where and how the company obtained the material it used to train its models.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Thinking Machines, the artificial intelligence startup founded a year ago by Mira Murati (former chief technology officer at OpenAI), unveiled its first open-weights general-purpose model, called Inkling, on July 15.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Important notice: the provided content of this Wall Street Journal article is only the initial fragment visible before the paywall. The text cuts off after the first paragraphs, so the summary that follows is strictly limited to that data and is necessarily brief; it is not possible to know…
🕒 Published on Zendoric: July 17, 2026 · 00:24
It was not possible to access the actual content of this article. What was downloaded from the Yahoo Finance link corresponds only to the site's cookie notice and privacy preferences, with no informative text about the news in question.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Apple Intelligence, Apple's generative artificial intelligence offering, has finally received approval to launch in China. According to Reuters, the Cyberspace Administration of China (CAC), the country's internet content regulator, gave the green light to Apple's AI services after…
🕒 Published on Zendoric: July 17, 2026 · 00:24
Masayoshi Son, founder and largest shareholder of SoftBank, has once again vehemently defended his bet on artificial intelligence during the company's annual corporate conference in Tokyo.
🕒 Published on Zendoric: July 17, 2026 · 00:24
Last Thursday, Mark Zuckerberg posted on X for the first time in three years. That fact alone, the author notes, already says something. The occasion was the launch of Muse Spark 1.1, the second model to come out of Meta Superintelligence Labs and the first Meta model to arrive with a price.