|
The Kitala Brief Welcome to the first Kitala Brief. You're one of the first on the list, and I'm glad you're here. Here's what this is, in one breath. Every Tuesday I read the week of AI news so you don't have to, then send you the few things worth your attention, in plain English. Some weeks that's a lot to cover. Most weeks it's a little, and I'll say so. No hype, no homework. Now, the first issue. Dave’s Read Before you let anyone tell you AI means a smaller team, sit with one finding. Gartner looked at large companies putting AI agents to work. Most had cut jobs. The ones that cut the most did not get better results than the ones that held steady. The layoffs freed up budget. They did not produce returns. (Gartner) What set the winners apart was plain. They redesigned how the work gets done. Their people got real training. And someone owned the result. So here is your thought for this week. Before you tie AI to a headcount decision, ask a smaller question. Where could AI give the team you already have two hours back? Start there. The savings are real, and you keep the people who know your business. The Few Tidbits That Matter OpenAI released GPT-5.6. The new models landed July 8. They are faster and steadier on long tasks. If your team uses ChatGPT, the update comes to you. You do not need to rebuild anything to feel it. Worth ten minutes to see what changes for the one or two tasks your team runs most. (CNBC) Leaders are getting surprised by the AI bill. Nearly half of companies scaled back an AI rollout this year after the cost came in higher than expected. Usage-based pricing hides the number until it lands. If you are paying for AI tools, ask a plain question: what did this cost per finished task, and was it worth it. (MarketingProfs) AI is quietly taking your website’s clicks. Google’s AI answers are cutting clicks through to websites, by close to 40% on some searches, and the visitors lost were often real prospects. The takeaway is old and still true. The audience you own, your email list and your direct relationships, is worth more every month the open web sends you less. (Search Engine Journal) The Rest You don’t need to read these. They’re here so you don’t feel like you missed anything. • OpenAI gave ChatGPT’s voice real-time, back-and-forth conversation. The fuller rundown, if you want it, is here. Seen in the Field A small, real result you can copy. Marketing teams that tested five or more versions of a LinkedIn ad, with AI drafting the variants, saw more than 20% higher click-through than teams running one or two. The ads are not the point. AI earns its keep when it helps you try several small versions of something quickly, then keep what works. (MarketingProfs) Worth Following Three people who explain AI without the noise. Ethan Mollick, Wharton professor and author of the One Useful Thing newsletter. The clearest voice for non-technical leaders on what AI can actually do. Allie K. Miller, AI advisor who writes plainly about putting AI to work in a business. Rowan Cheung, who runs The Rundown, a fast daily read on what happened in AI. (I'm a subscriber) For the Further Along If you are already running AI across a few tasks, here is one lever. Not every job needs the most expensive model. Summaries, sorting, and first drafts run fine on cheaper, faster models. Save the premium ones for the hard reasoning. Matching the task to the model is where a lot of teams quietly cut their AI bill without losing quality. Try This One skill you can build this week, in fifteen minutes. Open ChatGPT or Claude and paste this: “I run a [type] business with about [X] people. Here are the recurring tasks my team does every week: [list eight to ten]. Which three are the best candidates to speed up with AI, and what is the simplest first step for each?” You get a short, ranked list to react to. It will not be perfect. It will be a starting point, which is exactly the point. No fear, no hype, no FOMO. Just what’s worth your time. Reply and tell me what you are working through. I'd love to hear from you. Dave |
Every Thursday, the few things in AI that actually matter for your business, in plain English. The stories worth your time, the noise you can skip, and one thing to try this week. Just enough AI to be dangerous.
Issue 007 · October 1 to October 7, 2026 In this issue Tavus released an AI that holds a live video call, and 26 of the 54 people who took one thought it was a person. OpenAI began watermarking ChatGPT's writing, then published the trick that erases it. And a16z found almost a third of big companies claim AI results. About 2% track a number. Plus: ask your AI what it doesn't know about you. Dave's Read On October 1 Tavus released a video model where a video avatar carries on a live...
Issue 006 · September 24 to September 30, 2026 In this issue Meta released a free AI agent for small businesses, and it connects to QuickBooks, Stripe and Shopify. Microsoft rebuilt Copilot around agents that keep running after you log off, and NVIDIA started selling a way to box one in. The fifteen-minute job at the end is the one to actually do: three sentences that stop you repeating yourself to every tool you own. Dave's Read There's nothing to adopt in this one. Meta released a free AI...
Issue 005 · September 17 to September 23, 2026 In this issue ChatGPT now works inside Word, and a free account is enough. OpenAI is shutting off custom GPTs in December. Gemini got out of a test and into three real companies. Claude added documents and slides. Plus a question to ask around the office. Dave's Read For two years, using AI on a company document meant copying the text out and pasting it into a browser. That step was annoying. It was also the only thing between your files and...