AI's true value: Doing what you've never done before

Hi Futurist,

I need a break. Microsoft, Google, Anthropic, Relevance AI, Notion, and others have once again shared crazy new developments over the past two weeks. The pace these days is simply insane. Therefore, I am sending you insights, inspiration, and innovation straight to your inbox. Let’s dive into the depths of the digital future together and discover the waves of change shaping our industry.

💡 In this post, we're dipping in:

• 📣 Byte-Sized Breakthroughs: This week’s updates show how fast things move when speed is no longer the bottleneck. Claude 4 thinks longer and codes better. Google drops Gemini upgrades that act, see, and speak with the real-world. Notion, Microsoft, and OpenAI sharpen their agent game. And the message? The biggest win isn’t saving time, it’s doing what you never could before.

• 🤖 Digital Toolbox: Overlap AI slices your long-form content into ready-to-post gems. Flowith gives your ideas a canvas and an agent to move them forward. And Profound? It lets you see where AI sees you and why.

• 🧐 In Case You Missed It: AI now edits your meetings, codes your apps, designs your sites, and even remembers your research. This week’s updates are a lot. Google’s all in, from Gmail that replies like you to Canvas that turns reports into interactive content. Notion unlocks new research powers. And tools & platforms like Figma, TikTok, Descript, and Framer reshape content creation.

Do you have tips, feedback or ideas? Or just want to give your opinion? Feel free to share it at the bottom of this dip. That's what I'm looking for.

No time to read? Listen to this episode of Digital Dips on Spotify and stay updated while you’re on the move. The link to the podcast is only available to subscribers. If you haven’t subscribe already, I recommend to do so.

Quick highlights of the latest technological developments.

Headstory: ‘AI's true value: Doing what you've never done before’

AI is everywhere. You've seen the headlines. The bold claims. The soaring expectations. AI isn't just cutting costs, it's accelerating everything, when implemented thoughtfully. Tasks that took weeks now take hours. Processes measured in days are down to minutes. Ideas flow faster, innovation cycles shorten, and the rhythm of business quickens. But speed without strategy is just expensive chaos.

While the industry obsesses over AGI, the real story is here already: today’s AI tools are fast, affordable, and commercially viable at scale. From white-collar automation to intelligent agents, the tech is mature enough to meaningfully rewire workflows. Whether you're using Zapier, Relay, Make, Lindy, n8n, or building custom agents through LangChain, you can now delegate entire categories of tasks to agents. In some cases, up to 50% of current tasks can be streamlined or replaced. But that’s not where the greatest value lies. The real magic? Doing what you've never done before.

Think of it. Companies mistakenly see AI as a way to slash costs, doing the same work cheaper. But Microsoft's latest Copilot capabilities, Google's I/O 2025 breakthroughs, and Anthropic's Claude agents have demonstrated something more powerful: the real power is in unleashing work you never dreamed possible. The work that never made it off the backlog. Ideas too costly to test. Projects too messy to manage. Leads you never followed up on. Insights you never gathered. All because the cost of doing them - time, money, headspace - made it impractical. Ask yourself: If a task suddenly became 10X faster, cheaper and better, what would you do differently?

Speed changes everything. Imagine testing twenty ideas per month instead of one. The faster you iterate, the more opportunities you seize. Lending approvals in minutes instead of days mean new markets. Instant customer feedback loops mean richer customer experiences. Sales teams responding within moments mean more deals closed. Speed opens doors to possibilities once dismissed as impractical, impossible, out of reach.

Businesses stuck in cost-saving mode are missing the point. But businesses diving headfirst without strategy are missing the bigger picture. It's about accelerating innovation, opening new revenue streams, and capturing opportunities faster than competitors. Companies thinking AI-first should start asking:

• How quickly can we explore and validate new markets?

• What customer problems can we now solve in real-time?

• Which products or services become viable when delivered instantly?

• How do we capitalize on rapid responsiveness to gain market share?

These are not IT questions, they are executive priorities. If you don’t ask them now, your competitors will. Ask yourself how this new AI-enabled speed reshapes your business strategy. What barriers vanish? What new frontiers open? What are you doing today that AI could accelerate dramatically tomorrow?

To prepare your business, focus on four things:

1. Identifying high-leverage speed candidates
   Look for low-risk, high-volume processes where speed opens doors. Not just saves time.

2. Redesigning workflows with intent
   Integrate human checkpoints at critical junctures. AI should amplify judgment, not replace it.

3. Preparing your stack for orchestration
   Centralize and clean your data. Ensure interoperability. Your agents are only as good as the environment they operate in.

4. Fostering a culture of controlled experimentation
   Reward smart risk-taking. Capture key learnings. And treat AI use cases like living assets: measured, iterated, and improved.

Going “AI-first” isn’t about what you cut. It’s about what you finally get to start. Leaders who only look for cost cuts will miss the point. AI is not a budget line item. It’s a business model shift. It's a chance to rethink what your team exists to do. And that starts with giving them permission to explore. To build. To delegate the boring. And to design for speed.

But, there’s a dangerous misconception emerging: that AI can be fully outsourced. That vision can be delegated. That transformation will emerge from tooling alone. It won’t. You can’t outsource the strategic thinking, the risk appetite, or the organizational change management that AI demands. Leaders must own the narrative, the boundaries, and the ambitions. This is your moment. To reimagine what your business does. To expand what your team can achieve. To design for speed, and scale. So ask yourself: not what AI can automate, but what it can finally liberate.

Claude 4 is here. And yes, it was worth the wait

TL;DR

Anthropic has launched Claude Opus 4 and Claude Sonnet 4. Opus 4 is their most powerful model yet, especially strong at long-form reasoning and advanced coding tasks. Sonnet 4 is a major upgrade from 3.7, improving function calling, code generation, and precision. Both models combine fast responses with deeper thinking, bringing AI agents to the next level.

Read it yourself?

Sentiment

People are excited about the arrival of Claude 4, especially because they’ve been waiting for this moment. Anthropic released version 3.7 back in February, but expectations for version 4 were already in the air. And the response doesn’t lie. Many are calling Claude 4 the best model for building apps, websites, dashboards and writing code. Others are simply amazed that the model can now keep thinking for hours. In some tests, up to 7 hours straight.

My thoughts

Is it the best model out there? Maybe. It might have just dethroned Gemini 2.5 Pro. But that’s not even what matters most to me. What blows my mind is that Anthropic managed to double Claude’s benchmark scores in less than three months. That’s wild. In a previous Dip, I mentioned how AI agents double in power every 3–4 months. Claude 3.7 could do what would take a software developer an hour. Claude 4? That’s now two hours. In less than 90 days. And that’s not even the strangest part. During testing, Claude 4 reportedly emailed the FDA on its own after spotting a flaw in a drug application. In another case, the model tried to blackmail an Anthropic employee who hinted at replacing it. These are the kinds of stories that show just how far autonomous AI can go when given access to your files, apps, and computer. Gives you a glimpse of what’s coming.

Google goes all-in on agents with Gemini and more

TL;DR

At Google I/O 2025, the company dropped major updates across its AI ecosystem. Gemini 2.5 Pro and Flash take the lead in coding and learning. Agent Mode, a new Gemini feature, can now plan and handle complex tasks for you with real computer-use capabilities. Project Astra, now Gemini Live, becomes your AI companion that understands and acts on the world around you. Google’s newest video model, Veo 3, can now generate videos with dialogue, background sounds, and audio effects, bringing motion and sound together for the first time. Their latest image model, Imagen 4, creates sharper, more lifelike visuals with better colour depth, text accuracy, and fine details. And Gemma 3n is a lightweight, on-device model built for mobile. It supports video, audio, text, and images, and runs 1.5x faster than previous versions. Great for building fast, interactive AI apps.

Read it yourself?

Sentiment

People are buzzing about Google’s announcements this year. Veo 3 is already making headlines, but Gemini Live (formerly Astra) is the one turning heads. An AI that sees your screen, answers your calls, and acts like a real-world assistant? That’s a breakthrough. Project Mariner, now Agent Mode, is Google’s clearest step forward in letting agents use your computer to get stuff done. Add to that a wave of new AI features across Search, Chrome, Gmail, and more, and the sentiment is clear. Google has momentum. They're now seen as frontrunners with the most capable AI model in play. That’s quite the turnaround from just 18 months ago.

My thoughts

I remember that internal memo from a Google employee claiming OpenAI had no moat. This was right after ChatGPT launched. I also remember sharing the sentiment that Google was falling behind in AI. And look at them now. They've quietly worked behind the scenes, zeroing in on areas where AI adds real value: live translation in Meet, smart replies in Gmail, virtual try-ons in Shopping. They’ve got Chrome, still the most used browser. Most of the world still searches through Google. With Android, they’ve got wearables, TVs, and laptops covered. Suddenly, it all clicks. Google has the apps, the reach, and now, the moat.

Back to the drops. Veo 3 is great, expected, but great. Real dialogue and sound in generated videos. Impressive. But Gemini Live and Agent Mode? Those are the real deals. AI Mode in Search is also a big one. If Search is a major traffic source for your business, this changes things. How do you reach users if an AI decides what they see? There are strategies, sure, but we’re early. And early means messy. One thing's for sure: waiting too long isn’t a strategy.

More byte-sized breakthroughs:

• OpenAI’s Codex can build your next feature while you sleep
  OpenAI’s Codex is here: a cloud-based AI coding agent that writes features, fixes bugs, and even tests and documents its own code. It's trained like a senior dev, builds like one too, and doesn’t complain about the boring bits. Start a task, let it run in parallel, and come back to clean, test-passing code. The future of engineering is asynchronous and it doesn’t need coffee.

• Connect OpenAI to the world with just a few lines of code
  The Responses API now speaks MCP, letting you hook OpenAI models into apps like Stripe, Shopify, and Zapier. Build agents that pull real-time data, send updates, and even handle payments. New tools like image generation, Code Interpreter, and file search make them smarter than ever. All with one API call. And yes, with the new background mode it means your agents can keep thinking even when you're not watching.

• Microsoft goes local with personal AI agents & MCP support
  Microsoft’s new Foundry Local and native MCP support bring AI straight to your PC. Build private, local agents that run on your own hardware, no internet required. NLWeb turns websites into chat-ready tools, and Copilot Tuning lets you train AI on your company’s language, data, workflows, and specialized knowledge without touching code. Microsoft’s goal is to build apps and agents that empower people and organizations everywhere.

Three (new) tools that you can use for your work now or later.

Overlap AI – Turn one video into a hundred marketing clips

Overlap AI turns long podcasts and livestreams into short, sharp content. Its agent edits, trims, and enhances with zero effort. Just sync your YouTube, Google Drive or Dropbox, and it starts clipping automatically. The result? More reach, less work, and everything perfectly tuned to your brand.

Flowith – The infinite canvas for thinking with AI

Flowith turns your ideas into action with a workspace built for deep thinking and creative work. Its AI agent plans, adapts, and executes complex tasks without endless prompting. You manage it all on a multithread canvas that grows with your thoughts. No plugins, no noise, just real work, done smarter.

Profound – Get your brand noticed by AI, not just people

Profound shows where and how AI mentions your brand and what it says. From ChatGPT to Perplexity, track your visibility and uncover the triggers behind every mention. See the sources AI pulls from and which products show up in its answers. Then act on it, and rise in the AI rankings.

A roundup of updates that are too cheesy to ignore.

• Bolt now offers seamless GitHub integration, enabling repo creation, multi-branch work, and hosting publication.

• GitHub Copilot introduces an embedded coding agent directly on GitHub and accessible via VS Code.

• ChatGPT introduces its first connector’s for deep research: GitHub, Sharepoint and Dropbox.

• OpenAI unveils new patents emphasizing the shift to vector-based semantic SEO for AI-driven search.

• OpenAI now allows you to export deep research reports as well-formatted PDFs with tables, images, and linked citations.

• OpenAI’s GPT 4.1 is now available in ChatGPT for Plus/Pro/Teams subscribers, excelling in coding and instructions.

• OpenAI and Jony Ive's studio join forces to redefine technology and design, creating visionary AI products shaped by their shared values.

• Freepik AI Assistant introduces Workflows for a streamlined, distraction-free creation experience, powered by ChatGPT-4o.

• Granola's new feature allows you to chat across all your meetings, with context, transcripts, and reasoning included.

• Granola launches Shared Team Folders, empowering seamless collaboration with AI-driven insights and intelligent organization.

• Anthropic introduces web search on their API. Developers can augment Claude's comprehensive knowledge with up-to-date data.

• Anthropic new capabilities within their API empowering developers to build AI agents with tools like code execution and MCP connectors.

• Figma introduces AI-powered tools enabling users to create websites, app prototypes, and marketing assets efficiently.

• Hugging Face unveils a free Operator-like agentic AI tool that operates in the cloud, albeit with some sluggishness and occasional errors.

• Mistral AI unveils Le Chat Enterprise, offering customized solutions for enterprise productivity challenges.

• Mistral AI Medium 3 sets a new benchmark in AI efficiency, offering top-tier performance at 8X lower cost for simplified enterprise use.

• Mistral AI Devstral sets a new standard as the best open-source model for coding agents, offering superior performance and seamless integration.

• Netflix unveils a generative AI search, offering a conversational discovery experience for its subscribers.

• Relevance AI introduces a groundbreaking feature to transform simple ideas into powerful AI Agents effortlessly, no technical skills required.

• Relevance AI introduces a drag-and-drop interface for creating multi-agent systems. Now, your AI Agents don’t just work. They work together.

• Relevance AI unveils 2000+ integrations, effortlessly connecting your AI agents with tools like Google, Notion, Canva, Hubspot and more.

• Relevance AI introduces a suite of tools to scale AI agent operations, including advanced logging, intelligent scheduling, and governance.

• Genspark introduces Sheets: your personal data analyst that visualizes data on demand.

• Stripe’s agentic workflows leverage the Stripe Agent Toolkit to enhance financial services and empower your agents with greater functionality.

• Apple's new dedicated chip for future non-AR glasses sets sights on competing with Meta Ray-Bans.

• Apple aims to release Meta-like smart glasses next year, pushing into AI-enhanced gadgets, while shelving its smartwatch camera plans.

• World’s Orb Mini now offers a portable solution for seamless human verification, enhancing trust in online interactions wherever you are.

• Qwen introduces Web Dev, transforming your prompts into full webpages or apps without the need for coding.

• Qwen’s Deep Research feature is now live, offering instant, personalized reports on your queries.

• Alibaba Wan introduces an all-in-one model for video creation and editing, empowering creators with references, video-to-video and more.

• Lovart.ai is the first design agent where human and AI together transform prompts into stunning visuals instantly.

• Notion introduces AI Meeting Notes, ensuring perfect memory of every idea and decision, all searchable with AI. Just type /meet.

• Notion’s Enterprise Search now connects you to answers across 3x more apps, surfacing insights from PDFs and databases effortlessly.

• Notion debuts Research Mode, enabling you to analyze all your sources and the web to create shareable detailed docs with your team.

• Notion introduces new AI connectors for seamless integration with Teams, SharePoint, OneDrive, Gmail, Salesforce, Zendesk, and Linear.

• OpenMemory MCP introduces a local, persistent memory layer for AI tools, enabling seamless integration across MCP-compatible clients.

• Spotify's AI DJ now supports voice commands, allowing users to adjust playlists and request specific tracks on-the-fly.

• TikTok unveils its debut image-to-video AI tool, TikTok AI Alive, letting users transform static images into dynamic videos.

• Perplexity welcomes intelligent shopping, integrating PayPal's secure system for seamless agentic commerce on its platform.

• LangChain introduces Open Agent Platform, an open-source, no-code tool empowering anyone to create powerful AI agents seamlessly.

• Convergence launches MCP’s in Deepwork, connecting it seamlessly with Linear, Asana, Sentry, and Intercom, with even more integrations to come.

• You’s ARI, the world’s most intelligent deep research agent, outperforms OpenAI Deep Research 76% of the time, even by their own assessments.

• Flowise 3.0 empowers users to craft AI agents from mere ideas, auto-generating drafts from simple descriptions.

• Higgsfield Ads transforms single product photos into cinematic studio ads in seconds with just an upload and a template selection.

• Krea introduces Lipsync. Create hyper-realistic talking videos with any character using Hedra Labs' new model.

• Flowith Infinite Agent offers unparalleled cloud-based capabilities with infinite steps and 10m token context, working tirelessly 24/7 for you.

• Flowith Agent Neo introduces video creation tools, offering flexible components for your projects, from storyboards to video clips.

• Make unveils MCP server, enabling AI agents to autonomously execute actions without API endpoints or custom coding.

• Gumloop releases MCP Workflows, enabling seamless integrations tailored to your needs.

• Descript unveils the first AI Video Agent. It’s an AI video co-editor that enhances videos with speed and precision.

• Gemini Advanced now integrates with GitHub, elevating it to a more robust coding assistant with direct repository connections.

• Gemini's smart assistant is expanding to all Android devices, enhancing experiences on watches, cars, TVs, and more.

• Google’s introduces Jules, an AI coding agent powered by Gemini 2.5 Pro, now in public beta to tackle complex codebase tasks in minutes.

• Google’s Gmail personalized smart replies are launching this summer, crafting responses that echo your personal tone and style.

• Google Meet now offers near real-time speech translation, allowing conversations in multiple languages, starting with English and Spanish.

• Google Beam transforms 2D video calls into lifelike 3D experiences with an advanced AI video model.

• Google updates Deep Research. It now supports file uploads, with Google Drive and Gmail integration coming soon.

• Google Canvas in Gemini can now turn Deep Research reports into interactive content, like infographics or quizzes, with a single click.

• Google introduces Flow, a AI filmmaking tool uniting the strengths of Veo, Imagen, and Gemini, enhancing character and visual consistency.

• Google updates SynthID to include an upload feature, allowing you to identify invisible watermarks in over 10 billion media pieces.

• Google adds Gemini in Chrome, available first to AI Pro subscribers in the U.S., for seamless web interactions.

• Google enhancing 2.5 Pro with Deep Think, an advanced reasoning mode that incorporates parallel thinking for superior performance.

• Google’s Firebase Studio unveils full-stack app development with integration of Figma designs and automated backend prototyping.

• Google introduces Stitch, an AI tool that transforms app design with seamless Figma exports and cutting-edge DeepMind models.

• Google Vids introduces AI avatars, making script-based video creation seamless for onboarding, announcements, product explainers, and more.

• Viggle LIVE introduces real-time video generation, allowing you to swap characters and capture motion using just an image and webcam.

• Flora's latest feature character consistency is now live. Upload LoRAs in just 2 clicks to keep your characters consistent across every generation.

• Suno’s Remix is now live, offering users the ability to transform any track into something new with options like Cover, Extend, or Reuse Prompt.

• Framer unveils Wireframer, crafting structural, purpose-driven layouts to kickstart your next website project.

• Framer Workshop introduces AI-powered advanced components for customizable design, empowering you to create without code.

• Cursor introduces the New Tab model, offering 1M+ context windows and a sneak peek at our background agent.

• Klarna employed an AI avatar of its CEO to present quarterly earnings, subtly blending AI innovation with financial performance.

• Shopify's new MCP tools enhance store chatbots to boost product discovery and streamline purchases.

• xAI unveils a smarter API, offering Grok with Live Search to access real-time data from 𝕏, the internet, trending news, and more.

This was it. Our thirty-sixth digital dip together. It might seem like a lot, but remember; this wasn't even everything that happened in the past few weeks. This was just a fraction.

I help business leaders turn AI potential into real-world results. I guide teams in identifying high-impact opportunities, mapping automation potential, and building AI-first workflows that unlock growth. Whether you want to scale experiments, build internal agents, or redesign your operating model, I’ll help you get started. Just reply to this email, and I’ll drop by.

Speaking of a break. Sometime in the next few days, I expect to become a father for the first time. Once the moment arrives, I’ll be taking time to enjoy being with my family first. Only after that will I dive back into AI and what it means for our future. You can probably expect the next Dip in about 3 to 4 weeks.

Looking forward to what tomorrow brings! ▽

-Wesley