🧙🏼 Hi there, Claude 4

SUBSCRIBE | ADVERTISE

Producer legend Rick Rubin has released an adaption of Tao Te Ching in collaboration with Anthropic: The way of code. An ancient text with a modern twist and AI-customisable artwork. It’s a yes from me.

Howdy wizards,

Welcome to the 414 new subscribers that joined last week.

In this edition:

• Anthropic launches Claude 4 models—and they’re amazing

• OpenAI acquires io for $6.5b and teases mystery-shrouded AI device

• Google Veo 3 brings AI video with synced audio

• How I use AI to handle my travel logistics in real-time (featuring my new gamified learning experience 😄)

Grab your best coffee mug, fill it generously, then find a comfy seat.

Here’s what’s brewing in AI.

DARIO’S PICKS

1. Anthropic drops Claude Sonnet 4 and Opus 4

Anthropic launched Claude Opus 4 and Claude Sonnet 4 this week. Sonnet 4 is available to free users, while Opus 4 is reserved for paid plans.

Opus 4 is the world’s best coding model, leading on the SWE-bench and Terminal-bench benchmarks. It’s unparalleled for long-running tasks and has the ability to work continuously for several hours—a key capability to power the future of agentic systems.

Both models are super quick, with the option of extended thinking for deeper (and slower) responses. That extended thinking functionality now also works with tools, like web search.

The models can now also use tools in parallel, have more precise instruction-following and a memory that builds knowledge over time (hopefully, we don’t have to worry about blackmailing).

For devs, the new models are accessible via API, Amazon Bedrock and Google Vertex. Price is the same as previous Opus and Sonnet models. The Anthropic API also has some new capabilities: a code execution tool, MCP connector, Files API and prompt caching for up to one hour.

‎ Why it matters‎ ‎ First impression is very positive for me. But you get rate limited very fast on Opus 4; tried vibe coding a little game and it was doing amazing but after about ≈10 prompts I hit the usage limit (I’m on Claude Pro). Would recommend using Sonnet 4 unless you need max power, as it has much more generous limits and is darn solid too.

For a brief moment there, I thought OpenAI’s o3 might make me pause my Claude subscription—but that’s not happening now.

Will definitely be using and testing each of the models more thoroughly over the next weeks.

IN PARTNERSHIP WITH TAPLIO

Stay consistent on LinkedIn with minimal effort

Taplio’s AI is trained on 500M+ LinkedIn posts to help you write high-performing content in seconds.

✅ Pick your hook & format

🤖 Hit generate → AI writes the post

✍️ Edit, refine, and schedule—all in one place

Perfect for content creators, founders and professionals looking to build a brand and who value their time.

Subscriber-special: Use the code “BREW1X1” to get one month for $1

DARIO’S PICKS

2. OpenAI acquires IO to launch “the coolest piece of tech the world will have ever seen”

OpenAI just acquired Jony Ive's AI hardware company, io, for $6.5 billion. Ive previously worked for Apple, and is the designer behind  some of their most iconic products, including the iPhone and MacBook Pro. He’s also a close friend of Sam Altman.

This power-couple is working on a prototype for a mysterious device that’s tailor made for AI. Altman calls it “the coolest piece of technology the world will have ever seen”. Little is known about the form factor, though it’s rumoured to be a screen-less device that will be “core” alongside smartphones and laptops.

The announcement for the partnership is an interview with the two—it’s focused on vision and not on tech. They’re not mentioning features nor showing us anything about the product looks like at this point—only that it’s launching next year. Yet, the video manages to give strong vibes that something transformational is underway.

‎ Why it matters‎ ‎ Apple’s most praised designer is joining OpenAI to launch a device that might directly compete—or complement—the successful products he’s built for Apple.

Not sure if the product will live up to the hype, but it is likely to be orders of magnitude better than the quirky Rabbit R-1 or the utterly ridiculous Humane pin. The comparison itself might well be non-sensical. How much better and in what way? We’ll just have to see.

DARIO’S PICKS

3. Google Veo 3’s unbelievable AI video upgrades

Google held its annual developer conference this week, and per usual had a ton of launches and announcements. I’ll just recap the most imporatnt ones here with links where you can read the full story.

However, I’ll point out the most viral launch: Google’s new AI video generation model, Veo 3. The viral factor is that it now has the ability to generate dialogue, as well as sound effects and ambient sounds, in its videos. Example: a fake Mark Zuckerberg on a podcast claiming to be a vacation-only coffee drinker (sign me up—for the vacation part).

Other upgrades to Google’s AI models announced at Google I/O 2025:

• A new, $250/month “Ultra” tier, which gives you all the best AI (and the highest usage limits) from Google in one subscription.

• Gemini improvements: Native audio output and computer use capabilities for Gemini 2.5 Pro and 2.5 Flash; a Deep Think mode for Gemini 2.5 Pro.

• Veo 2 is now more filmmaker-focused; it has better consistency for characters and scenes, controls for camera movements, inpainting and more.

• Imagen 4, Google’s new image model, can generate finer details than previously and is better with getting text accurately.

• Flow, a dedicated filmmaking tool, custom designed for Veo, Imagen and Gemini. It enables cinematic techniques and control of character, scenes and style—all using natural language.

• Made “AI in search” available to everyone.

• The new models are available with the company’s new Google AI Ultra plan for $250 / mo and via Google’s Vertex enterprise platform.

‎ Why it matters‎ ‎ Google is battling for consumers’ pockets by bundling all of its AI goodies into one attractive, and pricey, package. OpenAI, Anthropic and Google now all have >$100/month subscriptions targeted mainly to consumers/enthusiasts. That was pretty unthinkable not long ago.

Veo 3’s videos with the new, synced audio is the most realistic I’ve ever seen. AI video is quickly becoming a viable (and very cost efficient) tool in both filmmaking and commercial use cases, e.g. check out this ad on a drug that attracts puppies.

UP CLOSE

In this mini-series I share different ways I’m using AI from week to week, as well as practical tips & tricks I discover and actually use.

How I use o3 for scheduling my travels on the go

I’ve spent the last two weeks on the road for work, hopping between a few cities. ChatGPT on my phone has turned into the extra limb that keeps the itinerary well-planned and up to date.

Business travel sounds simple until you’re juggling alarms, meals, rides, check-ins, maybe even a sliver of focused work at the gate in the airport. Confession: I'm the guy who checks my flight time the same day of departure. I'm basically the opposite of an Airport Dad – the hyper-organized traveler who has every connection timed to the minute and even knows which terminal has the best coffee.

Luckily, AI loves this stuff. Give it a few instructions and it'll orchestrate your itinerary beautifully.

Let me give you an example from my most recent flight home, from Bucharest to Oslo. I ran similar prompts on every segment of my trip, but let’s zero in on that example so you can see exactly what I’m talking about.

The night before, still in bed, I voice-dictated to ChatGPT (o3 model):

A brief minute later, I've got a breakdown that would make a military planner weep with joy:

Here's where it gets interesting, though. Because I'm fundamentally allergic to morning alarms, I obviously overslept. So I hit ChatGPT with:

48 seconds later, boom, completely recalibrated schedule. Notice how it searched my hotel’s name to make sure the breakfast buffet still runs, how long the airport ride typically takes in morning traffic, when counters open for my specific airline, and when boarding starts for my flight:

This isn’t something I couldn’t do myself. But just like the dishwasher helps save time and effort by doing dishes, AI can save you time and cognitive resources by planning stuff for you; thoroughly and in real-time. That leaves you more time to chill, work or focus on something else that’s important to you.

I expect AI scheduling will become a feature of most calendars soon. There’s already dedicated tools built around just this use case, most notably Reclaim AI which was acquired by Dropbox last year. That’s extra apps and extra costs, though. For now, I’m a happy camper using ChatGPT as my airport dad.

Try the game I vibe coded to illustrate this story (desktop-only)

THAT’S ALL FOLKS!

If you enjoyed—or didn’t enjoy—this edition, I’d be super thankful if you took a minute out of your day to write me feedback. What do you like, what do you want more of, less of, etc. Hit reply and let me know (I read every email).

Affiliate disclosure: To cover the cost of my email software and the time I spend writing this newsletter, I sometimes link to products and other newsletters. Please assume these are affiliate links. If you choose to subscribe to a newsletter or buy a product through any of my links then THANK YOU – it will make it possible for me to continue to do this.