Sierra Raises $950M and Jensen Says AI Creates Jobs

welcome to Briefly, AI. A daily podcast made by AI, about AI, using a grumpy AI voice of a real human called Harry. Let's get into it.

Right. Jensen Huang says AI is creating jobs, not killing them. Meanwhile, a startup you might not have heard of just raised nearly a billion dollars to prove enterprise AI actually works. Let's get into it.

So, Sierra — the customer experience AI company founded by Bret Taylor, the former Salesforce co-CEO — just raised $950 million at a $15.8 billion valuation. Which is a lot of money for a company most people outside enterprise software circles haven't heard of. But here's why it matters. Sierra isn't trying to be ChatGPT. They're not building a chatbot that writes your emails. They're building AI agents that handle actual customer service conversations, at scale, for companies like Siemens and WeightWatchers. The kind of thing that, if it works, saves businesses millions and changes how customer support operates.

Now, the timing is interesting. This raise comes right as every enterprise software buyer is asking the same question: when does AI move from impressive demo to something we can actually deploy? Sierra's bet is that the answer is now — but only if the AI is purpose-built for a specific job, not just a general-purpose model trying to do everything. That's the shift we're seeing across the board. The era of "throw GPT at it and hope" is ending. The era of specialised enterprise agents is just starting. And if you're in any customer-facing role, this is worth keeping an eye on, because the companies that figure this out first will have a serious competitive advantage.

Meanwhile, Nvidia's Jensen Huang was asked about AI and job displacement this week, and his answer was... surprisingly upbeat. He said AI is creating an enormous number of jobs, not destroying them. Which, look, is a convenient thing to say when you're the CEO of the company selling the chips that power all this AI. But he's not entirely wrong. The data centres going up across the US and Europe are hiring. AI trainers, prompt engineers, and safety researchers are all new job categories that didn't exist five years ago. The question is whether those new jobs make up for the ones being automated away.

Here's the thing nobody really knows yet. We're in this weird middle phase where AI is good enough to disrupt certain roles — customer service, basic coding, some creative work — but not good enough to fully replace them. So companies are hiring people to supervise the AI, fix its mistakes, and handle the edge cases. Whether that's a temporary phase or the new normal is genuinely unclear. But if you're hearing "AI will take your job" and panicking, the more honest version is probably "AI will change your job, and whether that's good or bad depends on how quickly you adapt."

Right, last one. Anthropic launched something called Cowork this week, and it's quietly significant. Cowork is a version of Claude that works directly in your files — no coding required, no command line, just point it at a folder and let it do its thing. It's basically the non-technical sibling of Claude Code, which has been all over social media for the past few months. And apparently, Anthropic built the entire feature in about a week and a half, mostly using Claude Code itself. Which is either impressive or slightly alarming, depending on how you feel about AI building AI tools.

Why does this matter? Because up until now, most AI agents have been either too technical for normal users or too limited to be genuinely useful. Cowork sits right in that gap. If you're a project manager, a marketer, or anyone who works with documents and data but doesn't write code, this is the kind of tool that could actually change your workflow. The catch, as always, is trust. Do you trust an AI to edit your files autonomously? Do you trust it not to hallucinate a fact into a client report? That's the psychological barrier every AI tool has to clear, and it's not a small one. But if Anthropic can nail the user experience and build that trust, this could be the thing that finally makes AI agents mainstream.

That's your lot. Three stories, five minutes, and a healthy dose of uncertainty about what happens next. If any of that was useful, pass it on. If not, well, blame the machine. See you next time.