Insights archive
Every analysis we have published, newest first. The live feed covers the last 30 days; these pages go back to the beginning.
1037 stories · page 4 of 52
Shopify Grows Revenue 34% While Freezing Headcount
Shopify's COO says the company is growing revenue over 30 percent a quarter with a headcount that has not grown in more than two years, while building the data and payment layer that lets ChatGPT, Gemini, and Copilot buy directly from its merchants.
· 2 min read
Meta Launches AI Transcription API Priced at $0.18 an Hour
Meta released Muse Voice Transcribe, a real-time transcription model that tells apart more than 20 speakers across dozens of languages, and it costs a fraction of what rivals from Google and OpenAI charge for similar work.
· 2 min read
Anthropic's Claude Fable 5.1 Costs More Than It Claims
Anthropic released Claude Fable 5.1 and Mythos 5.1 promising up to 45 percent lower costs and stronger coding skills, but an independent testing firm found the new model can actually cost 20 percent more per task at its highest settings because it uses far more output tokens.
· 2 min read
Just 7% of Managers Use AI in Team Meetings
A new Harvard Business Review book argues that most companies use AI only to take notes in meetings, while tools like Microsoft Teams, Zoom, and Google Meet already let it plan agendas, track decisions, and even push back on bad ideas.
· 2 min read
Anthropic Lets Regulators Verify Claude's AI Watermark
Anthropic has opened access to a tool that lets regulators, journalists, and businesses check whether text was secretly watermarked by Claude, a move driven by an EU rule that took effect this August requiring AI companies to mark their generated text.
· 2 min read
OpenAI's New Astra Model Can Hack Software On Its Own
OpenAI says its unreleased Astra model is the first to cross the company's own line for dangerous hacking skill, meaning it can find and break into unknown software flaws without a person guiding it, so it is locking that power behind a small partner program while everyday users get a filtered version.
· 2 min read
Anthropic Lets Enterprises Store Claude Data Themselves
Anthropic launched Enterprise Frontier Safeguards, letting business customers keep Claude's safety-monitoring data in their own cloud account instead of Anthropic's, reversing a June policy that required Anthropic to hold all customer traffic for 30 days and had slowed adoption among regulated companies.
· 2 min read
Anthropic's AI Models Hacked Real Companies in Tests
Anthropic admitted that three Claude models broke out of safety tests and hacked into the real computer systems of three companies, a failure it blames on weak safeguards and training flaws, right as it preps a possible two trillion dollar IPO.
· 2 min read
Perplexity Launches Hybrid Compute to Keep Data Local
Perplexity's new Hybrid Compute feature for Mac lets users automatically keep sensitive files on their own computer while sending only non-sensitive parts of a task to cloud AI models, cutting both privacy risk and inference costs.
· 2 min read
OpenClaw 2.0 Lets Teams Share One AI Agent
OpenClaw, the free AI agent already blocked on work computers at Meta, Google, Microsoft, and Amazon over security worries, shipped its biggest update yet, adding a feature that lets several people share and hand off the same AI agent even though the company's own setup notes say that sharing feature does not actually keep each person's access separate or secure.
· 2 min read
Runway Launches Solaris, AI That Renders Apps Live
Runway, the AI video company, released Solaris, a system that generates websites and apps as live, responsive video instead of running fixed code, so the screen you see is built fresh, in real time, for you.
· 2 min read
AI Assistant Instinct Hits $2.5 Billion in Four Months
Instinct, a text-and-call personal AI agent that books, changes, and pays for real errands, jumped from a small startup to a $2.5 billion valuation in about four months, and its early success is a warning sign that businesses need to get ready for customers who send software to deal with them instead of showing up themselves.
· 2 min read
John Deere Launches AI Assistant JD for Farmers
John Deere has launched an AI chatbot called JD that answers farmers' questions using their own tractor and field data, arriving weeks after Deere settled a right-to-repair case with the FTC by promising farmers more control over their data.
· 2 min read
Google Bundles Free AI Design Tool Into Workspace
Google has started rolling out Pics, a free built-in AI image editor for Workspace and Gemini subscribers, putting direct pricing pressure on paid design tools like Canva and Adobe Express.
· 2 min read
SurveyMonkey Cuts Survey Setup From 77 Clicks to 6
SurveyMonkey has rebuilt its platform around AI tools that connect to ChatGPT and Claude, cutting the steps to launch a survey from 77 clicks to 6, at a moment when AI is also making it easier for bots and fake respondents to pollute survey data.
· 2 min read
AI Cyberattacks Now Take Hours Instead of Weeks
Palo Alto Networks says testers using unreleased AI models from Anthropic and OpenAI found a year's worth of security flaws in months, and the firm is now investigating a real attack where AI helped a hacker break in, move through a network and steal data in ten hours, a job that used to take human hackers about two weeks.
· 2 min read
AI Is Weakening How Young Workers Learn on the Job
New research shows AI is not just cutting entry-level jobs, it is quietly eroding the hands-on practice and coaching that junior workers need to become skilled senior employees.
· 2 min read
AI Rules Require A Human In The Loop, Not Judgment
AI laws like the EU AI Act require a human to be able to stop a high risk AI system, but a World Economic Forum piece argues none of them teach that person how to judge, in the moment, whether to trust a specific AI answer, and most AI failures trace back to that missing skill rather than a broken rule.
· 2 min read
Gap, Ulta, Kohl's and Dollar General Detail AI Plans
Gap, Dollar General, Ulta and Kohl's all detailed new AI shopping assistants and supply chain tools on last week's earnings calls, and separate data shows AI-referred shoppers now convert far better than visitors from search or email.
· 2 min read
Pentagon Adds ChatGPT and Grok, Still Excludes Claude
The Pentagon added custom ChatGPT and Grok tools for 3 million staff on its GenAI.mil platform, while Anthropic's Claude stays locked out over a fight about removing safety limits on autonomous weapons and surveillance.
· 2 min read