In partnership with
Exploring AI Voice With SuperBloom
Scaling a campaign globally means finding a voice that resonates everywhere—without losing the overall message. For Deel's "Feeling of Deeling" campaign, agency SuperBloom needed exactly that: a consistent brand voice across markets, deployed fast, without sacrificing quality or consent. They built it with Branded AI Voice, powered by real, professional talent.
SuperBloom and Voices break down how the campaign came to life in this on-demand video session—from strategic talent selection through seamless production workflows and global scale, to the governance that future-proofs an audio strategy built to last. You'll hear directly from the team on how they made this happen, plus their advice if you're looking to explore AI voice for your brand.
If you're a marketing executive, agency creative, or brand leader mapping campaigns in international markets, watch this on-demand session to get a real playbook, not a hypothetical—lessons any creative team can apply to their own global rollout.
For almost two years, we've been talking about AI agents that would manage our schedules, handle errands, and keep working while we did something else.
There have been useful tools and real progress along the way. Lately, though, the idea is starting to feel much closer to everyday life.
I bought a new Mac mini to test personal AI agents.
Projects like OpenClaw and Hermes let you set up an assistant with memory, tools, and ongoing responsibilities. A dedicated computer gives it somewhere to work. You can also use an existing machine or a cloud server; the model itself may still run remotely. With a setup you manage, keeping that environment running is part of the job. OpenClaw, Hermes
Now Muse and Grok Bot provide cloud computers for their agents. You connect your tools, set permissions, and hand over work that can continue after you close your laptop. Muse, Grok Bot
I really like how easy Muse is to use. Grok Bot has also been useful, although it felt a little scrappy at first.
I bought hardware to explore this. Many people will get their first useful experience by sending a message.
The signal
Some of the most interesting examples are coming from people putting these agents to work in their own lives.
He put Muse in the producer's chair. ThursdAI host Alex Volkov used Muse to follow his live show and update the on-screen headlines as the conversation moved. He also gave Muse and Instinct the same dinner-ticket task. He reported that Muse had booked his Rosh Hashanah tickets while Instinct was still looking for the RSVP. An AI producing your show and sorting out your evening plans starts to feel pretty personal. Alex's account
He gave Grok Bot a process. It created four more bots. Jamie McKean says he handed it his company's 50-plus-step website process. Within ten minutes, it had created bots for preparation, building, quality checks, and technical setup. Two hours in, he had a website ready for human review, plus a bookkeeping bot reading Xero and Stripe, supported by a separate CFO bot. His WhatsApp experiment also got his account blocked. Useful, ambitious, and still rough around the edges. Jamie's original post
His agent negotiated $4,200 off a car. Then he bought it a Mac mini. In January, AJ Stuyvenberg used OpenClaw, then called Clawdbot, to contact three Hyundai dealers and send competing offers back and forth. He reports a $4,200 dealer discount. He kept approval over consequential replies and took over for the paperwork. Afterward, he bought a Mac mini to give the agent its own place to run. I can relate to that last part. His X post · His full account
That January experiment predates OpenClaw 2.0, released August 30. The update simplified setup, rebuilt the browser interface, and introduced shared sessions so another person can join ongoing work with the context intact. The release
Instinct makes the interaction feel like an ordinary text. An Atlantic writer gave it $25 to choose a book for his editor. He reported that it researched the editor's writing, selected a book, and arranged local pickup. The reported test
A live show produced. A website built for review. Dinner tickets booked.
These are the kinds of results that can turn curiosity into a habit.
My read
The shift is from believing personal agents will matter to deciding what we are ready to trust them with.
For me, ease of use is a meaningful part of that shift. If getting help requires managing another complicated system, the benefit can disappear quickly.
The agent that earns a place in someone's life will need to remember useful context, follow through, and ask sensible questions before taking consequential actions.
My bet is that people will care less about which agent gives the most impressive first response and more about which one they still want working for them a month later.
Put it to work
Give one agent a small recurring responsibility this week: preparing for meetings, watching a deadline, or producing a morning briefing.
Define the result, connect only the sources it needs, and say when it must ask permission.
After five days, check three things: Was it useful? Was it accurate? Did it save more attention than it demanded?
Repeated usefulness is how an agent earns more responsibility.
The human side
I am excited about what these tools can take off our plates. I also know how easily saved time becomes room for another project.
I want these experiments to create more space for meaningful work and the people in my life.
An assistant should make that easier. Otherwise, I have bought myself another thing to manage.
The question
What would a personal AI agent need to do for you to keep using it?
Reply and tell me.
Until next week,
Burhan
P.S. Ready to put your agent to work? I’m building Awesome AI Agent Skills, a free, open-source collection of 103 skills for tools like Claude Code, Codex, and Cursor. It’s already nearing 200 GitHub stars. Pick a skill, try it on a real task, and if you find it useful, give the project a star to support it and help more people discover it.