What the rapid rise of Muse, Instinct and Grok Bot means for business
Over the last two months, the AI industry has quietly tipped from "agents are coming" to "agents are here." Meta launched Muse on September 8, and it has already been downloaded an estimated 3.4 million times, at one point briefly overtaking ChatGPT as the most-downloaded free app in the U.S. App Store. Instinct, the SMS-based personal agent, has crossed 100,000 users while still invite-only, and its valuation rocketed from roughly $500 million in early August to $10 billion by September. That’s a 20X jump in under two months, without the company disclosing a single growth metric to justify it. xAI's Grok Bot, aimed more at professionals and enterprises, launched in August and expanded from a narrow beta to full enterprise access in just three weeks. Three very different companies, three very different approaches, and all of them converging on the same bet: that the agent is about to become the default way people get things done online.
The interface is changing, and it's changing fast
That's the real story underneath the download charts and funding rounds. Meta's Muse is putting genuinely capable, multi-step agents in front of everyday consumers in a way that earlier tools, like OpenClaw, never managed to do at scale. Muse isn't a novelty chatbot bolted onto an app; it's a real attempt at an agent that can carry out tasks on someone's behalf, and the download numbers show it's landing with ordinary users, not just early adopters.
Instinct took a different, maybe smarter, path to the same goal. It turned SMS text messaging into an AI agent interface, with no dedicated app at all. Nobody has to learn a new UI pattern or download anything. Everyone already knows how to text. By meeting people inside a channel they've used for twenty years, Instinct sidestepped the adoption curve that trips up most new software. I’ve been using this for the past week and it’s very impressive. It doesn’t feel like AI. It feels like you are texting with a really smart, and hardworking personal assistant.
Even people who've been skeptical of chatbot-driven interfaces are starting to concede the ground is shifting. In a recent TechCrunch interview, Airbnb CEO Brian Chesky, who has long been skeptical about agents, talked about having personally used Airbnb through both Muse and Instinct. His verdict: It doesn't work very well…Yet. In his words, hotels are rough but Airbnbs work even worse. But notice what he's not arguing. He's not saying agents are a dead end; he's saying the industry hasn't yet built the infrastructure for agents to do their jobs well and that apps need real developer kits so agents can plug into them properly instead of fumbling through screens built for humans. That's not a rejection of the agentic future. That's an admission that it's arriving faster than most companies, Airbnb included, are ready for. This is something that most brands need to be thinking about, their own interaction with agents.
There are still questions about privacy and security that have come up. Meta’s use of data has come under question for years and it has surfaced again with Muse. Not everyone is comfortable with how these agents work and it will cause some friction in adoption. However, just like social media, the utility of these tools outweighs these issues for customers.
Why open wins: less friction
This brings us to a quieter story playing out alongside the download numbers: Amazon's decision to wall off Muse from its full product catalog. On the surface that looks like Amazon protecting its turf. In practice, it's an opening. Every brand and every shopper Amazon decides not to serve through its own agent is a customer that Shopify, Walmart, or the next platform can pick up instead.
Here's the thing most of the agent conversation misses: the consumer doesn't care which rail fulfills the order. They don't care whose agent found the flight, booked the hotel, or reordered the detergent. They care about getting what they want, quickly, without friction. The moment one company makes that harder by gating access or forcing a walled-garden experience, it hands a gift to every competitor willing to stay open. Shopify and Walmart don't need to out-build Amazon's agent. They just need to be the path of least resistance when Amazon isn't. In a market moving this fast, openness isn't a philosophy, it's a growth strategy.
None of it matters without the data
Now for the part of this conversation that gets the least attention and matters the most: none of it works without the data underneath it. It doesn't matter whether the interface is a chat window, a text thread, or a cloud agent signing into your tools. It doesn't matter which model is doing the reasoning. If the data feeding that agent is incomplete or ungoverned, the output will be mediocre no matter how slick the front end looks.
Our recent AI Maturity survey found that 69% of companies investing in AI are still stuck exploring, experimenting, or implementing. They are not actually scaling. That's not a model problem. That's a readiness problem. Only 29% of organizations have completed a formal data audit, and just 5% believe their data is actually AI-ready. This one bears repeating, just 5% believe their data is ready. Confidence is outrunning governance, too. A quarter of leaders admit they can't fully explain or trust their own AI outputs, 36% say they lack internal AI expertise, and only 3% have a documented AI-audit process. Customers also want to know how their data is used and protected. This is a consideration to get ahead of to gain and retain consumers' trust when it comes to adopting and working with these agentic tools.
How to adapt your brand
Put these threads together and the picture is clear. The agent wave is real, and the download and funding numbers prove consumers are willing to meet it where it already lives, whether that's a dedicated app or a text thread.
3 things to consider:
- Think and test how your brand interacts with agents. Try the experience yourself like Brian Cheskey did. Is your brand ready? Does the experience match your expectations, or what you deliver via your website?
- Openness will beat walled gardens, because customers are more attracted to reducing friction, than brand loyalty. Accept that agents are going to be a way consumers interact with your brand. Think through your technology, pricing, and UX with agents.
- The winners in this next phase won't just be the ones with the best agent. They'll be the ones whose data can actually support one.
At Further, we have been helping brands prepare for, navigate, and succeed in the AI space. We do this using strategy, governance, and creating strong data foundations. Read our AI Maturity study to learn more about how we are leading enterprise AI adoption and helping businesses confidently navigate the future.
