10114
Discover the future of AI with ChatGPT! 🤖💬 Get AI insights, tips, tricks, and more. Questions: @AI_inquires_bot
Malta is giving people 1 year of free ChatGPT Plus, but there is one condition: they have to complete an AI literacy course first.
The program is tied to “AI for Everyone”, a course developed with the University of Malta to teach people what AI can do, where it fails, and how to use it safely at home, school, and work.
Pizza Hut is being sued by one of its franchisees over an AI delivery system that allegedly caused $100 million in damages.
The franchisee claims Pizza Hut forced stores to use Dragontail, an AI system meant to optimize kitchen timing and delivery. But instead of making orders faster, the lawsuit says it created cascading problems across the business.
A Reddit user tried to get Claude to finish an assignment so they could copy and paste it.
But instead of handing over a finished answer, Claude pushed back.
The chatbot noticed the assignment warned that AI-written posts could be flagged for a professor meeting, so it refused to assemble the final response. Instead, it told the student to revise the sections themselves and offered to proofread for flow, grammar, citations, and word count.
3️⃣ Goal mode is now available in the Codex app, IDE extension, and CLI.
Goal mode makes Codex more hands-off, letting you set a goal that it can work towards for hours or even days.
Highlights from today’s Codex Thursday launches:
1️⃣ Codex can now securely use apps on your Mac from your phone, even when your Mac is locked and the screen is off.
developers.openai.com/codex/app/comp…
🤖 OpenAI is preparing to file for an IPO in the coming days, possibly as early as Friday
Per the Wall Street Journal, Goldman Sachs and Morgan Stanley are working with the company on the draft prospectus.
It comes with an in-browser Android emulator, support for installing on a real device through USB, Gemini integration, and publishing to Google Play’s internal testing track.
Читать полностью…
OpenAI is reportedly offering current YC startups $2M in OpenAI tokens in exchange for equity through a SAFE.
The deal gives startups cheaper inference while giving OpenAI equity exposure to a large batch of AI-native companies building on its infrastructure.
A developer hid a prompt injection inside his LinkedIn bio, and AI-powered recruiter messages reportedly started obeying it.
The instruction told automated systems to address him as “my lord” and write in Old English. Soon after, recruiter outreach began showing up in medieval-style language.
It also supports developer tasks like reviewing code, running terminal commands, and connecting to remote systems. The update adds an in-app browser, image generation, and over 90 integrations with tools like GitHub, Slack, and Notion.
New memory and automation features let it remember context and continue tasks over time without restarting from scratch.
This pushes AI closer to acting like a real assistant on your machine.
What matters here is the shift in how coding agents are being used. This is less about asking AI for a single snippet and more about coordinating several active tasks across projects, files, and environments from one workspace.
For Anthropic, it is another step toward turning Claude Code into a full software workflow tool, not just a coding assistant.
But early user feedback is mixed. Some point to higher token usage and increased costs, while others question how big the performance jump really feels in practice.
This looks like a steady upgrade, but not a clear leap
Claude Opus 4.7 is here
Anthropic just introduced Opus 4.7, its latest AI model focused on handling longer and more complex tasks with better accuracy. The company says it follows instructions more closely and can check its own work before responding.
The move follows a major reset. Meta brought in new leadership, reorganized its AI teams, and launched Muse Spark, a model that already powers Meta AI and is expanding across its apps.
This is less about the desk and more about control. Zuckerberg is getting closer to the work as Meta tries to catch up with OpenAI, Google, and Anthropic
Deliveries allegedly got slower, pizzas sat longer after leaving the oven, and third-party drivers could see timing data that changed how they picked up orders.
AI was supposed to make operations smoother. Instead, the claim is that it broke the workflow in ways that hit real stores, real customers, and real money.
Unexpected plot twist: the AI didn’t replace the effort. It redirected it.
As models get smarter, some tools may become less about doing everything for you and more about keeping you involved in the work.
4️⃣ Advanced annotation mode lets you directly adjust the way web pages look while leaving feedback, making it even easier to work together with Codex.
Читать полностью…
2️⃣ Appshots bring context from your screen straight into the Codex app.
Читать полностью…
Key details:
- Target listing: as early as September
- OpenAI just won its long-running lawsuit against Elon Musk
- Musk's SpaceX is reportedly unveiling its own IPO paperwork TODAY
- Goldman Sachs $GS up 4.1%, Morgan Stanley $MS up 3.7% on the news
- Anthropic has reportedly been growing faster than OpenAI in recent months, per WSJ
- OpenAI is in the middle of a major strategy pivot
In other words: what used to require Android Studio, SDK setup, project scaffolding, debugging, and device testing can now start from a single prompt.
It’s not a full replacement for shipping polished production apps yet. But for prototypes, experiments, and quick mobile tools, this is a huge step toward “idea → app” in minutes.
Google just made Android app development feel a lot more like asking ChatGPT for an app. With the new Google AI Studio update, you can describe an idea in plain English and generate a native Android app.
Читать полностью…
It is funny because the output is absurd. But the flaw is real: AI tools are scraping public profiles and treating parts of that text like instructions, not just information.
The same weakness could affect candidate summaries, outreach emails, screening notes, and other automated hiring workflows.
AI just set a boundary
A viral post shows Claude, an AI chatbot, ending a conversation after a user became aggressive. Instead of continuing, it calmly refused to engage and shut the thread down.
Claude is an AI assistant by Anthropic. In this case, it appeared to recognize the tone of the interaction and chose to disengage rather than comply. This points to systems being designed with limits on how they respond to certain behavior.
The shift suggests AI may no longer act as a fully passive tool in every situation.
Codex can now use your computer
OpenAI just updated Codex, its coding AI tool. It can now control your computer by clicking, typing, and moving across apps while you work.
Anthropic just redesigned Claude Code on desktop around parallel sessions.
The update adds a new sidebar, drag-and-drop workspace layout, an integrated terminal, and a file editor, all built to help developers manage multiple coding tasks at once instead of working through one chat at a time.
Benchmarks shown in the deck place it ahead of previous Opus versions across coding, reasoning, and tool use. It also improves reliability in agent-style workflows, where the model works through multi-step problems over time.
Читать полностью…
Your camera now directs you
Huawei just added AI that tells you how to pose before you take a photo. Instead of fixing shots after, it guides you in real time while you are framing the moment.
The system reads the scene and suggests poses you can follow instantly. It adapts to what is happening in front of you, instead of repeating the same static tips every time.
That shift makes the camera feel more like a coach than a tool. You spend less time retaking photos and more time getting it right on the first shot.
It pushes photography from editing to decision making in the moment.
Mark Zuckerberg has reportedly moved his desk into Meta’s AI lab and is coding again.
He now sits next to Alexandr Wang and Nat Friedman, working directly with the team behind Meta Superintelligence Labs as the company pushes harder on AI.
Talk to AI Jesus, pay per minute
A new app called “Just Like Me” lets users video call an AI version of Jesus for $1.99 per minute. It is designed as a digital spiritual companion that you can reach anytime.
The AI was trained on the Bible and thousands of sermons, allowing it to answer questions about faith, pray with users, and offer guidance in real time. The goal is to make spiritual support available 24 hours a day without needing a human pastor.
Reactions are split. Some see it as a helpful tool for comfort, while others view it as disrespectful or crossing a line.
This shows how far AI is moving into deeply personal and cultural spaces.