The challenge with AI: navigating the ambiguity of tools and applications

Imagine a world where AI is being used to reconcile high conviction transactions, where you can ask a bot to give you a summary of your financials, and fetch guidance from the ATO within your accounting platform #XeroCon presented a view of that world across Xero and the apps that integrate As we transition from synthesising (generative) AI to actioning (agentic) AI, we're starting to see interesting use cases developing for how emergent technology can be applied in a practical way While what was previewed is still in development, it's an early perspective of a world where automation of business is a given, and richer data can be presented faster to business owners. Exciting times ahead

The 6 levels of AI prompting. Most firms are still stuck at level 1. That’s why they only see AI as a shiny toy. But there’s an entire curve of value waiting to be unlocked. 1️⃣ Zero-shot - quick answers to technical questions 2️⃣ Few-shot - categorisation & payroll adjustments with context 3️⃣ Chain of thought - step-by-step breakdowns of complex calculations 4️⃣ RAG - pulling live data from Xero or HMRC 5️⃣ Fine-tuned - models trained on your own client history 6️⃣ Agents - autonomous systems monitoring cash flow and flagging risks The higher up the curve you go, the more transformative the impact. And here’s the best part: you don’t need a lab or millions in budget. You just need the awareness (and the will) to start climbing. Where does your firm sit today? If you’re curious about how firms are really using AI today, that’s exactly the kind of story I share in The Naked Accountant. // 𝗦𝘂𝗯𝘀𝗰𝗿𝗶𝗯𝗲 𝘁𝗼 𝗧𝗵𝗲 𝗡𝗮𝗸𝗲𝗱 𝗔𝗰𝗰𝗼𝘂𝗻𝘁𝗮𝗻𝘁 Real lessons on building, growing, and exiting a modern firm. It’s authentic, unfiltered, and real. And it’s 100% written by humans ❤️ Go to my profile and click 𝗩𝗶𝗲𝘄 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿 to subscribe. Or visit https://naked.ac/subscribe

Great read from Troy Tempest Offer on how companies can get moving quickly with AI by looking inside first vs finding the shiny new external AI toy. #AIInnovation #PEInsights #gotomarket #RevenueGrowth

"If your role is to set GTM standards across the portfolio, this is now your lane." Troy Tempest Offer is spot on in this piece. Great advice for those of us rolling AI into our work. #AIInnovation #PEInsights #gotomarket #RevenueGrowth #RevOps

When I started at Cortado Group just over 18 months ago, we were building our AI solutions through workflow and automation tools such as Make and Retool. Along came Agentic coding tools such as Open AI’s Codex and all of a sudden, as a small team, we are able to rapidly develop standalone applications that do (within reason) anything we want. No longer are we bound to the limitations of capability that workflow systems will inevitably have, as they are generalist solutions, not specialist. Nor do you need a large software development team to build solutions, you only need a small number of key talented individuals.

Troy Tempest Offer said "The mandate isn’t to adopt AI. It’s to use AI to reduce vendor dependency, free up talent, and improve operating leverage. The companies that figure this out first don’t just grow faster. " Great statement on how AI (or any technical solution) should focus on the outcome. #ai #pe #privateequity

Troy Tempest Offer said "The mandate isn’t to adopt AI. It’s to use AI to reduce vendor dependency, free up talent, and improve operating leverage. The companies that figure this out first don’t just grow faster. " Great statement on how AI (or any technical solution) should focus on the outcome. #ai #pe #privateequity

Everyone’s chasing AI. But not enough teams are thinking about the data it runs on. The biggest takeaway from INBOUND this year wasn’t another AI feature. It was Data Hub. HubSpot’s been evolving in this direction for a while, but this was the first time it felt real. You can now bring in data from spreadsheets, events, apps—clean it, structure it, and actually use it. Not just for reporting. For real workflows. That’s what makes AI useful at scale. Not prompts. Not plugins. Structure. If your data’s a mess, AI can’t help you. But if your foundation’s solid, Data Hub turns HubSpot into something way more powerful.

AI is not magic. Buying a subscription won’t transform your business overnight. Real impact comes when teams dig in, experiment, and use AI to automate and simplify the way work gets done; one project at a time. Read our latest article on why AI business automation is about practice, not hype. https://lnkd.in/eJpKYci3 #AI #ArtificialIntelligence #AIinBusiness #BusinessAutomation #AIAdoption #DigitalTransformation #FutureOfWork #AutomationStrategy #EnterpriseAI #AIInnovation #ProcessAutomation #BusinessEfficiency #AItools #BusinessGrowth #TechStrategy #WorkplaceInnovation #AIforBusiness #AIimplementation #BusinessLeaders #AIAutomation

Nobody actually knows what they’re doing with AI. That should make you feel better. Lots of folks are posting big AI wins. Case studies. Playbooks. Frameworks. What they’re not showing is the part where they’re still guessing. That’s the truth: most people are still figuring this out. The tech’s changing daily. No one has it all dialed.  It’s all pretty kludgy. So if you feel behind, you’re not. You’re just early. Curiosity wins right now. Try things. See what sticks. Toss what doesn’t. In the day-to-day, we're using it to assist with a bunch of things, but they're mostly small or separate use cases.  Content creation, E-commerce analytics, brand messaging, financial reporting, proposals, content creation, prototyping, deck creation, and the list goes on. AI hasn't replaced anyone at F+L. There’s still a human in the loop all the time, and we expect it to be that way for the foreseeable future.  AI just allows us to do more with what we have, to be better at what we do.  What’s one AI use case you actually rely on?