Riley

the easiest way to stand out with ai right now: use more of the features in the tool you already have. problem is, it's hard to know what's even in there. so i made a tool to help. 1. think of a task 2. pick your tool (chatgpt, claude, copilot, gemini, grok) 3. answer 5 quick questions 4. see which features fit it best chasingnext.com/ai-feature-f…

close second: when they go build something on their own and show you what they made. feels like you gave them their wings.

unpopular opinion: ai task forces are doing more harm than good. talking with teams, i regularly hear that: - a few people are driving their ai strategy - they are stretched thin (and mostly not in ai roles) - the rest of the team is left waiting for direction i know teams have commonly turned to task forces to solve big problems, but ai is a different beast. there's no "right way" to use it. ai's value is specific to each person and corner of a business. if you want to capture its potential, everyone needs to be part of figuring out how. leaving them out causes strategy to suffer and culture too. according to @charterworks workers who have a say in how they use ai get the most out of it. but only 37% of employees say they do, compared to 67% of managers. so what should teams do? start from the ground up. first, give everyone the support to figure it out. get them access to tools and hands-on training that helps them get going. then let them find where ai fits in their work. this is how bigger opportunities surface, and when task forces can actually help.

seems counterintuitive, but if you want to get good at ai, learn to set up your own automations in a tool like n8n, make, or zapier (even though ai can do a lot of it for you now). it teaches you to: - break work into steps - see where ai fits in a process and where it doesn't - map "if this, then that" rules and paths - get specific on actions/results needed at each step - predict edge cases and troubleshoot bugs - adjust prompts until you get the result you want - decide when things run and how results are delivered - know what tool access each step needs newer ai can build all of this for you (and you should absolutely let it), but knowing how to set it up yourself lets you build stronger systems because you can steer ai, instead of letting it run on assumptions.

the best part about helping people implement ai is being the one who gets to blow their mind by showing what it can do for them.

the thing that surprised me most about teaching ai: how much people get from just watching me work. if you're helping someone learn, let them watch you: - turn off training settings so ai's not using private data - open new windows when fresh context makes sense - use multiple windows and bounce between tasks - have ai help write prompts for better results - schedule tasks and make things reusable - connect tools and adjust permissions - push back on answers and steer ai i used to think most of this was wasted time in sessions, but this is the part clients like best.

a lot of ai resistance is just people who haven't figured out how it helps them yet.

signs a team is getting good at ai: 1. people bring up how ai can help without being asked 2. teammates asking to see what others built 3. ai now feels energizing instead of feared 4. there are more ideas than time

the fastest way i've found to get better at ai: watch what others are building and pull the good pieces into your own setup. this is different than using someone's full build, which usually falls flat since it was made for their work. what i recommend: - pick where you'll look (x is the place for this right now) - follow people who share what they're building - send anything interesting to ai as you find it - ask "what pieces could i apply to my work from this?" - tell it what ideas to pull out when you already know - customize what you take to fit your work - have it keep a list of ideas you don't have time for yet - have it push back to avoid overbuilding

what orgs hiring for ai enablement actually ask for: - drive ai adoption across teams - run training and workshops - lead change management - build a champions network - define ai fluency levels by role - write the curriculum - pick and prioritize use cases - map and redesign workflows - run pilots - build agents and automated workflows - set governance and guardrails - write playbooks and templates - teach people how to check ai's output - set up evals for ai output - run hackathons and demo days - share success stories internally - brief executives - partner with every function - know copilot and claude - measure ai's impact

i've been helping a client figure out how to use ai for their job, meanwhile their company has only given them access to the basic copilot plan. it's made one thing clear: a lot of orgs expect their teams to transform with ai, but are only giving them the most limited versions of tools. with copilot basic, you can: - use chat with basic web search - save prompts you reuse (if you can find them again) - set up notebooks that reference uploaded files - ask about your email/calendar in outlook but you can't: - ask about email/calendar in copilot chat - schedule prompts to run on their own - connect agents to sharepoint or onedrive - use copilot in word, excel, or ppt at larger orgs - use researcher or analyst agents a ton of people get stuck prompting because the features that let them build workflows to reuse are locked. it's not just this client... i've heard of other orgs giving advanced access to select people and holding everyone else back too. how are people supposed to change how they work when tools they need are locked?

what % of grok bot users do you think are also x users?

i've been setting up ai for a small business owner this week. tool-wise, i recommended they get both claude and grok bot. would've loved to recommend only one tool, but grok bot is too helpful to skip. a few things working in grok's favor: - low friction for someone new to ai use - ability to create multiple focused bots - ability for bots to work with each other - easy to get it to take action and work on routines - low maintenance and adapts easily to feedback - scraping information from gated sites - sees bigger picture across chats/connectors well - easy to use on the go here's how they'll use grok bot vs claude: grok bot (day-to-day & on-the-go work): - inbox triage: sorting + what needs attention - customer service: suggested responses, management - pulling information: access data claude can't - general support: shopify and business questions claude (complex tasks, data aggregation & analysis) - audits: site and shopify reviews and suggestions - web scrapes: high-volume reports and analysis - data processing: digging for trends and recos - context: gathering and curating info for grok bot - routines: set up structured work w/ skills a few more thoughts - both can work together with shared google drive files - claude duet isn't a replacement for grok bot yet - claude too helpful for complex work to leave out - think grok bot will be their fav (simplicity wins)

looking for backlinks for your site? submit it to vibecuterie.com only takes a min :)

the more i talk to people, the more i realize the average person has no idea what ai can already do. for most people being good at ai means chatting with it.

having terminal copy prompts for other windows to my clipboard instead of fumbling in docs >>>

crazy time we're in. if you release an innovative digital product, you can bet someone else copies it in a week.