Why Voice AI Is Really a Trust Solution for Small BusinessHere’s a thought that should stop every small business owner mid-scroll: the technology you’re most afraid of might be the one that finally levels the playing field. Not because it replaces you — but because it protects your most valuable asset. Trust. That’s the core argument Laurent, a French-born entrepreneur who splits his time between the US and Israel and has worked in digital since 2002, made on a recent episode of The Special Marcoting Live Show. Laurent is the force behind GetOblic, a voice AI platform built specifically for small and local businesses, and his perspective on AI is refreshingly human.Most conversations about AI fall into two camps: breathless hype or existential dread. Laurent refuses both. He spends roughly twelve hours a day working with AI — not as the core of his business, but as a constellation of assistants — and his conclusion is blunt. AI will not replace humans. It will replace some tasks, sure. Some jobs, yes. But not us. Why? Because AI without human judgment is a liability. And judgment, as he puts it, is a purely human quality.Judgment Is the Skill AI Can’t AutomateLaurent doesn’t sugarcoat the limitations of AI. He’s caught AI lying — not in a subtle, interpretive way, but blatantly fabricating information. When he called it out, the AI admitted it. His takeaway isn’t panic. It’s perspective. Five years ago, if a human lied to you, you’d either exercise judgment or be naive. The same rules apply now.The scalability problem compounds this. Ask AI one good question, and you’ll probably get a great answer. Ask a derivative of that question twenty times, and the quality starts to erode. Push for five hundred or a thousand operational items, and the whole thing falls apart. AI, as it stands, doesn’t scale the way we want it to — and Laurent suspects there’s a financial reason behind that. Platforms don’t want to spend too many tokens, so they try to dump as much information as possible into a single response. For the person on the other end, trying to build something step by step, that’s a real workflow problem.The antidote? Diligence, organisation, and a framework that Laurent’s team uses daily.The Operational Book: A Framework for AI ContinuityOne of the most practical takeaways from the conversation is what Laurent calls the Operational Book, or Op Book. The problem is simple: when you’re deep into a project with AI, you know when you start but never when you finish. You need sleep. The AI doesn’t. And when you pick things up the next morning — often in a new thread because the old one is overloaded and sluggish — you’ve lost context.The Op Book solves this. Every time Laurent and his team hit a meaningful threshold in their work, they ask the AI to write a comprehensive summary: every task completed, every problem encountered, every solution found. Then they download it as a PDF and store it locally. The next day, when they open a fresh thread, they upload the Op Book to give the AI a running start.It’s not bulletproof, Laurent admits, but it helps tremendously. And there’s a deeper strategic reason to do this. What happens if ChatGPT disappears tomorrow? What if you want to migrate from Claude to Gemini in six months? If your operational knowledge lives only inside a chat thread on a single platform, you’re one outage away from losing everything. The Op Book is insurance — platform-agnostic, portable, and human-controlled.Trust Is the Product, Not the TechnologyThis is where Laurent’s thinking gets genuinely distinctive. When his team at GetOblic set out to build a voice AI solution, they assumed they were building an AI product. They quickly realised they were building a trust solution.The reasoning is sharp. Small business owners spend years building relationships with their customers — often through their own voice, in person or over the phone. That relationship is built on trust. Now, if you delegate that trust to a voice AI and don’t disclose it, you’re breaking the very thing that made your business work.Laurent goes further. GetOblic can technically clone a business owner’s voice. They refuse to do it. Because imagine a long-time customer calling in, hearing what sounds like the owner, and three minutes in, something feels off. That’s not just an awkward interaction — it’s a lie. And a lie, in Laurent’s framework, is a trust break you don’t recover from.This conviction runs through everything they do. Their social media uses AI-generated video — and they leave the watermarks on deliberately. They don’t care. The generation isn’t the point. The content and the message are the point. Disclosure isn’t a concession; it’s the strategy.Laurent extends this principle beyond his own company. He flags how easy it is to spot AI-generated text now — the dashes, the spacing, the bullet point patterns, the telltale phrasing....
続きを読む
一部表示
1 時間 15 分