Everything in <30 Seconds

News, Tools & Resources

🧠 Claude 4 Is Here

Anthropic releases Claude 4, featuring enhanced reasoning, vision, and steerability.
Read the release → Anthropic blog

🎨 Jony Ive Joins OpenAI

Apple’s legendary designer takes on a new role shaping OpenAI’s future—think design meets superintelligence.
WSJ coverage → Read the article

🤖 ByteDance Drops open-source AI agent

An open-source, multimodal AI agent with vision-language capabilities—built to rival GPT-style interfaces.
See the details → MarkTechPost

🚀 Google I/O 2025: AI Everywhere

Major AI announcements from Google’s I/O 2025:

• Gemini 2.5 Pro: Advanced reasoning with 'Deep Think' mode.

• Project Astra: A universal AI assistant with real-time interactions.

• Veo 3 & Flow: Next-gen video creation tools with realistic audio-visual synthesis.

• AI Mode in Search: Enhanced search experience with AI overviews.

• Android XR Glasses: Smart glasses integrating Gemini AI.

• Stitch: AI-powered app design and development tool.

• SynthID Detector: Tool to identify AI-generated content.

• Google Beam: 3D video calling platform.

Full recap → Google Blog

🛍️ Shopify’s AI Horizons

From AI image generation to smart inventory predictions—Shopify showcases what's next in e-commerce AI.
Read the blog → Shopify Blog

🎧 Lenny’s Podcast: Microsoft’s CPO on AI Prototyping

“If you aren’t prototyping with AI, you’re doing it wrong.” Microsoft CPO shares bold takes on AI, product, and the future of building — in conversation with Lenny Rachitsky.
Listen on YouTube → Watch now

🌍 Big Picture: Work Is Still Trapped in the Past

We’re living through one of the biggest shifts in productivity since the internet — and yet most companies are still operating with an old school blueprint.

The mistake?

Most leaders are treating AI as a bolt-on to legacy job roles, not a chance to rethink how work actually happens.

Instead of asking “How can AI make this role faster?”, they should be asking:

• What are all the underlying tasks in this role?

• Which ones can be automated, augmented, or elevated?

• How can I rewire my team to orchestrate across people + machines?

📌 MIT calls this approach Deconstruct → Redeploy → Reconstruct which reminds me of my Deconstruct & Rebuild approach I once wrote about here.

That’s the new productivity loop.

AI frees people from busywork — but only if you restructure around tasks and outcomes, not rigid roles.

As AI democratizes skills, the “specialized experience gap” narrows. Less specialized people can now operate at a senior level — if they know how to use the tools.

🧬 Rethinking the Startup Playbook

The classic startup playbook used to be as follows:

• Raise big

• Hire specialists

• Layer in management

• Chase growth, ignore burn

• Raise until you have scale to make it profitable

Not anymore.

The most forward-thinking companies are small by design. They don't scale people — they scale leverage.

What’s different:

• No silos

• No middle managers

• No ticketing systems

• Everyone is a builder and owner

These are tiny teams built around versatile generalists — not deep specialists.

And it works because the toolset has changed. Anyone who knows how to operate AI can go full-stack.

⚙️ Case Study: Inside Gamma

Gamma is an example of what “the future of work” looks like in practice.

In a post on X, their CEO XZY explained how they flipped the old school model entirely by exclusively hiring "player coaches" who both lead AND execute.

• The old way: Find seasoned managers with impressive titles who direct others' work.

• The new way: Find leaders who explicitly say, "I still love doing the work myself."

They take a two-track approach by (i) identifying hidden talent within and (ii) implementing trial periods for new hires before both sides commit.

But their top signal? The absence of hubris.

Gamma is looking for people that show a "no job too small" mentality and a willingness to roll up their sleeves alongside the team.

Their hiring model:

• No traditional interviews

• 3-month paid trials

• Only hire “player-coaches” — people who lead and execute

• No job is “too small” for the CEO or Head of Design

Their culture:

• People build across domains

• AI is baked into everyone’s workflow

• Speed > structure

• Curiosity > credentials

🚀 What AI enabled work looks like

🧑‍💻 Growth PM Builds Self-Served Analytics Dashboard

• Doesn’t file a ticket to data science / engineering team

• Uses Claude to draft SQL

• Spins up a Retool dashboard

• Shares across team — no dependency, no delay

  → Insight in days, not weeks

🧠 Marketer Uses 1,000+ Chats to identify Personas

• Doesn’t wait for UX team to analyze and create personas

• Dumps user convos into Claude

• Prompts: “Summarize key frustrations. Build personas. Segment by use case.”

• Uses insights to reshape strategy and content

  → Instant research. No UX team needed.

🎨 Designer Prototypes, Tests with Users, and Ships

• Doesn’t wait for engineering and UX team

• Picks up signal in Slack

• Builds flows in Figma

• Prototypes using bolt.new, cursor.com, or v0.dev 

• Tests with users via Typeform + GPT

  → Idea to live test in under a week

The best generalist don’t just “do more with less.”

They’re operating on a different system of work.

Here’s a great cheat sheet created by Chris Donnelly

📌 Takeaways

AI is great at going deep. Generalists are great at navigating across.

They:

• Synthesize info across fields

• Operate without guardrails

• Frame the right problems

• Build loops, not just deliverables

• Communicate, adapt, lead

And in a world where tools like Claude, Cursor, Midjourney, and NotebookLM remove friction, the speed to execution is often what matters most.

This isn’t about knowing everything.

It’s about knowing how to figure anything out — fast.

If you're building a company:

• Design around tasks, not titles

• Hire for range, curiosity, and execution

• Replace interviews with trials

If you’re navigating your career:

• Stack skills: product, data, storytelling, ops

• Use AI tools like an exoskeleton — they’re your leverage

• Make your curiosity your most valuable asset

If you're leading a team:

• Break up silos

• Let people “opt in” where they can create value

• Redesign around outcomes, not ownership

Final Thoughts

There’s never been a more exciting time to build and improve your output. Take advantage of the tools and resources and deconstruct your work so you can strategically rebuild it and become more productive.

Until next time — keep building, keep exploring, and rock on.

-Andrea