- Everyday AI
- Posts
- Ep 807: Claude Tag, Gemini Home, Copilot Excel Skills and 7 other AI Features Available Today you Should Be Using
Ep 807: Claude Tag, Gemini Home, Copilot Excel Skills and 7 other AI Features Available Today you Should Be Using
The White House is urging OpenAI to delay its next model, Codex is now in ChatGPT mobile, and OpenAI may postpone its IPO until next year.
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Outsmart The Future
Today in Everyday AI
8 minute read
🎙 Daily Podcast Episode: This week brought a wave of new AI features that can help you automate work, collaborate better, and get more done today. Give today’s show a watch/read/listen.
🕵️♂️ Fresh Finds: OpenAI may let users gift Codex credits, Google is adding free AI study notebooks and practice tests to Gemini, and Runway has launched Agent 2.0 for marketing campaigns and more. Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: The White House is urging OpenAI to delay its next model, Codex is now in ChatGPT mobile, and OpenAI may postpone its IPO until next year. Read on for Byte Sized News.
💪 Leverage AI: Here’s the AI updates you mighta missed and how to use them today. Keep reading for that!
↩️ Don’t miss out: Miss our last newsletter? We covered: Report: White House thought Anthropic’s CEO was ‘weirdo’ in Fable discussions, Apple raising prices as AI chip demand drives up costs, Anthropic accuses Alibaba of model theft and more. Check it here!
Ep 807: Claude Tag, Gemini Home, Copilot Excel Skills and 7 other AI Features Available Today you Should Be Using
All frontier AI models are reportedly delayed. ⏳
New reports show that frontier models there were supposed to be released this month like GPT-5.6 and Gemini 3.5 Pro have been delayed as a result of the ongoing Anthropic vs. U.S. Government dispute.
So while we may have to wait a few more weeks for SOTA AI models, we DID get a ton of new AI features and model updates that are available now.
Tune in as we dish the 7 new AI features that can change your workflow today.
Also on the pod today:
• GPT 5.5 Instant gets chattier 🤖
• Google Finance AI tools launched 📊
• Claude Tag: Slack’s AI manager 💬
Listen on our site:
Subscribe and listen on your favorite podcast platform
Listen on:
Here’s our favorite AI finds from across the web:
New AI Tool Spotlight – Note.md is A local-first research workspace for Mac, Cewso is One AI for chat, images, voice, and live web search, Zaro Builds the tools your team actually needs.
OpenAI Giftable Credits — OpenAI looks to be adding giftable Codex credits, so you could soon send someone compute instead of a gift card.
Copilot Finance — Microsoft is pushing Copilot in Excel deeper into finance, with skills, trusted data connectors, and cleaner audit trails built for real workflows.
Gemini Study Notebooks — Google is rolling out free study notebooks in the Gemini app, with custom lessons, quizzes, and progress tracking built around your materials.
Gemini Practice Tests — Google is adding full-length ACT and GRE practice tests to Gemini for free, with Princeton Review content behind them.
How AI Transforms Work — Codex is no longer just for coding, it’s becoming OpenAI’s default AI tool across teams, including Legal, Finance, and Recruiting
Google Finance Updates — Google Finance is out of beta with global portfolio tracking, custom market briefings, and a new Android app.
Runway Agent 2.0 — Runway just introduced Agent 2.0, a tool that turns a simple prompt into full marketing briefs and campaign assets.
1. OpenAI Model Release Faces White House Delay Push 🫸
The Trump administration has reportedly asked OpenAI to stagger the release of its next model over security concerns, putting a federal spotlight on how quickly advanced AI should reach the public.
According to The Information, the request signals growing unease in Washington about the risks tied to releasing more powerful systems too fast, even as the AI race keeps moving at full speed.
2. Codex Hits ChatGPT Mobile 📲
OpenAI says Codex in the ChatGPT mobile app is now generally available, marking a new push to let developers manage coding work from their phones while the agent keeps running on connected devices like laptops, Mac minis, or devboxes.
The update adds one-to-one device pairing for more secure connections, plus notifications, goals, side chat, file previews, and inline review comments, which makes the mobile setup feel much less like a remote stub and more like a workable command center.
3. OpenAI May Delay IPO Amid Choppy AI Markets 📉
According to the New York Times, OpenAI is now leaning toward delaying its public offering until next year, with advisors urging CEO Sam Altman to wait as recent AI stock volatility spooks the market.
The timing matters because OpenAI had confidentially filed its IPO plans on June 8, but the path to a blockbuster valuation looks less certain after SpaceX and Cerebras both saw sharp post-IPO reversals.
4. Qualcomm jumps on Meta data center CPU deal 🖥️
Qualcomm shares surged after the company said it struck a multi-generation agreement with Meta to supply data center CPUs for Meta’s next server fleet, a timely move that puts the chipmaker deeper into AI infrastructure.
The first chip, the Dragonfly C1000, is aimed at high-efficiency computing at scale and is set to begin production in the second half of 2028, while the broader deal signals a longer-term relationship beyond one product.
5. Micron briefly tops Meta in market value on AI demand 📈
Micron Technology briefly surged past Meta Platforms in market value and even nipped at Tesla’s heels after a strong forecast sent its shares jumping Thursday, underscoring how fast AI-linked chip stocks can move when investors get excited.
The memory maker said customers have committed $22 billion to secure chip supply, a sign that demand is running hot and future sales look locked in. Micron’s rally pushed its valuation to about $1.398 trillion, while Meta stood at $1.392 trillion and Tesla at roughly $1.4 trillion.
6. Study: 25% use Codex for tasks estimated to take humans 8+ hours
A new study from OpenAI, Columbia, Duke, and the University of Pennsylvania says Codex use is surging, with OpenAI employees generating 99.8% of their output tokens through it and outside organizations now at about 17% adoption, up from near zero in August 2025.
Among individual users, 80.6% made at least one request estimated to take more than 30 minutes of human work, 70.2% had at least one that could save over an hour, and 25.6% delegated work estimated to take more than eight hours.
Most enterprise teams are still debating which AI tool to adopt. This week, the biggest labs skipped that conversation entirely and just moved AI inside the platforms everyone already uses.
The gap stopped being about models. It's about whether AI is already woven into how your team actually works.
Claude Tag turned Slack channels into AI-managed workspaces where anyone can delegate tasks to a shared Claude identity. Google shipped a $99 Gemini-powered smart speaker for the home.
Microsoft brought reusable AI Skills directly into Copilot for Excel, and Canva collapsed the entire ad lifecycle into one AI-native flow with Grow 2.0.
Sooo much shipped this week.
Your competitors prolly have at least two of these running by Monday morning.
OpenAI just shipped a new version of GPT-5.5 Instant, and this time the focus is on making ChatGPT's default model more fun to talk to. Better intent recognition, more reliable handling of complex constraints, and tighter conversational flow overall.
This is the third update since May 5. That's unusual for a model serving one billion weekly active users.
Paid users already have access, and free users should be getting it today or over the weekend. OpenAI rarely name-drops the Instant model more than once per release cycle, which hints this could be closer to GPT-5.6 performance wrapped in GPT-5.5 Instant clothing.
If your company runs on ChatGPT, the default experience just got measurably sharper without anyone having to change a single setting.
Try This
Start a conversation in Instant mode for quick back-and-forth work, then switch mid-conversation to a thinking or pro model when you need deeper analysis.
That context carries forward, giving you speed and depth in one session without starting over.
2. Google Finance Exits Beta With AI Tools 📊
Google Finance just exited beta and launched a dedicated Android app with AI-powered portfolio tracking, a Gemini-driven research tool, and a feature called key moments that explains why individual stocks moved. Upload a screenshot, CSV, or PDF of your holdings and start asking questions about your portfolio composition.
Scheduled briefings run in the background. Describe a task in plain language, like a daily premarket crypto analysis, and Google Finance handles it automatically without any follow-up from you.
Portfolio and task features are live on the web now, with the AI research tool available in the Android app today and an iOS version rolling out later this year. It's free.
A Gemini-powered financial research assistant at zero cost puts serious pressure on every paid analytics platform that still hasn't shipped AI features of its own.
Try This
Head to Google Finance on the web, upload a screenshot or CSV of your current holdings, and ask the research tool which sectors are underrepresented in your portfolio so you can spot allocation blind spots in under two minutes.
One query replaces the manual portfolio review that most people never actually finish.
3. Google's Gemini Smart Speaker Is Here 🔊
Google just launched the Google Home Speaker, a $99 device and the first audio hardware built from the ground up for the Gemini for Home ecosystem. The key difference: natural multi-step conversations where you can ask follow-ups without repeating yourself.
This gap has been wide open for years. Alexa Plus has disappointed users, and Apple's Siri faces lawsuits over failing to deliver on the AI voice assistant promises it made.
Available now across the US, Canada, UK, much of Europe, Japan, Australia, and New Zealand. Google Calendar, Gmail, and Shopping integrations let you schedule meetings, triage email, and shop by voice with real conversational context carrying through.
Smart speakers just went from timers to real assistants, and the first company to normalize one in every employee's home wins a brand new surface for daily AI adoption.
Try This
Pick one up, connect it to your Google Calendar, and ask it what next Tuesday looks like, then add a meeting just by describing the details out loud to see how much faster voice-first scheduling feels compared to pulling out your phone and typing it all in manually.
You reclaim every minute you currently waste on tiny-screen calendar entries.
4. Claude Tag Brings AI Teammates to Slack 🤖
Anthropic just launched Claude Tag, a new way for teams to work with Claude directly inside Slack by granting it access to channels, tools, data, and even codebases. The big difference from previous Slack AI integrations is that this is multiplayer, with one shared Claude identity per channel that everyone can see, steer, and hand off to.
Think of it less like a personal assistant and more like a channel manager that remembers everything and never drops a thread.
Available now in beta for Claude Enterprise and Team plan customers, replacing the existing Claude in Slack app with a 30-day admin migration window. Toggle on the optional ambient mode and Claude proactively surfaces information and follows up on stalled threads without anyone asking.
Anthropic reports that 65% of its product team's code now runs through an internal version of Claude Tag. That says a lot about where collaborative AI work is heading.
Try This
If your org is on a Claude Team or Enterprise plan, add Claude to one department channel and connect it to the tools that team relies on most.
Assign a real task, like chasing down a data question or triaging support tickets, and let Claude work asynchronously while the team moves on to other priorities to see how quickly the channel adopts an always-on AI teammate.
5. Canva Grow 2.0 Automates the Ad Lifecycle 📢
Canva just rolled out Canva Grow 2.0, expanding its performance marketing platform to automate ad creation, publishing, and performance optimization in a single AI-native workflow. It generates static and video ads using your brand context, audience signals, and past performance data.
The real unlock is bulk publishing across LinkedIn, TikTok, and Meta simultaneously, with creative that refreshes automatically based on what's actually driving results in each channel.
Available now for Canva users, building on the original Canva Grow from October 2025. Canva unveiled it at Cannes Lions this week, signaling a serious push to position itself as a full marketing automation platform rather than just a design tool.
Teams still stitching together three or four separate tools for creation, distribution, and optimization just got a compelling reason to consolidate, especially now that the creative can refresh itself without anyone manually swapping out underperforming assets.
Try This
Pull your top-performing brand assets into Canva Grow 2.0 and set up a campaign across two platforms to let it generate size-specific creative variations automatically.
Review the performance dashboard after one week and double down on whatever creative is actually working best across channels.
6. Gemini in Chrome Adds Select From Screen 🔍
Google just shipped Select from screen in Gemini for Chrome, a new tool that lets you highlight any text or image on your current browser tab and add it directly to a Gemini prompt. No more screenshotting, copying, or describing what you're looking at.
Just click, drag, and ask. It's that simple.
The feature runs on the computer use capabilities Google just built into Gemini 3.5 Flash, and it works right inside the Ask Gemini sidebar without leaving Chrome.
Rolling out now in Chrome 149. Make sure your browser is updated.
This shipped alongside the Gemini 3.5 Flash computer use developer release for building custom agents that can see and act across browser, mobile, and desktop environments, but the two features are technically separate.
Anyone who spends time manually feeding on-screen context into AI prompts just got a shortcut that collapses the whole process into a single click and drag.
Try This
Update Chrome to version 149, open the Ask Gemini sidebar, and use Select from screen on a complex chart or dense data table you would normally screenshot and describe by hand.
Notice how much sharper the response lands when the model sees exactly what you see.
7. Copilot Skills Make Repeatable Excel Work Disappear 📗
Microsoft just launched reusable Copilot Skills for Excel, letting teams automate common finance processes like building a DCF, closing books, refreshing monthly reporting models, or preparing variance analysis. A skill guides Copilot through each step, applying the right structure and formatting so the output is easier to review, reuse, and trust.
Custom skills use an open-standard markdown file saved to OneDrive. That makes them shareable across teams.
You can even import workflows already built in other AI tools like Claude or ChatGPT.
Prebuilt finance skills are available now for Microsoft 365 Copilot customers across Excel for web, Windows, and Mac. Custom skills are in the Insider channel today, with general availability rolling out next month.
If your finance team still rebuilds the same analysis every single month by hand, this is the feature that turns that entire workflow into a reusable one-click operation that any team member can run without rebuilding the logic from scratch.
Try This
Open Copilot in Excel, go to the Add Work Content menu, and select All Skills to browse the prebuilt finance templates available right now.
Run one on a recurring monthly task and compare the structured output to what your team produces manually.






Reply