🧭 THIS WEEK AT AI SECOND ACT
Howdy, changing up the theme of the website & logo a bit this week.Hope you like it. Thinking this week about the value of AI vs company limitations.
My goal is to make this as valuable and practical as possible as we navigate the new AI era. 🚀
🧰 AI NEWS + LEARNING
Here are a few things I found recently:
Claude Imagine - New tool from Claude, you may want to try out
One year of agentic AI - Lessons learned
Loneliness - AI could amplify this - and it’s already bad
OpenAI Sora 2 - Impressive new video generation from OpenAI
ChatGPT Pulse - The start of ‘personal assistant’ but only premium tier currently
Climate impact of GenAI - AI climate impact is growing considerably - A huge topic to consider and solve - AI must solve more than the impact!
🗺️ FEATURED INSIGHT
Have you tried internal AI tools vs commercially available tools? Internal chat vs OpenAI ChatGPT or Anthropic Claude?
I’ve noticed internal tools can be limited vs commercially available tools.
Chat with the internal tool - can you upload information [files, knowledge] and get the right answers?
Does ChatGPT or Claude or Gemini exceed the answer or knowledge from the internal tool?
Do you not have access to Copilot for Office? The organisation trying to save $20 per month per head? What?
This is a huge problem for slow, old, boring, conservative companies.
The investment spent on internal tools that perform worst than commercially available tools is a massive waste of money
The $$ withheld for tools that obviously can enhance productivity is the 2nd huge waste
In short, don’t accept internal tools that are not performing. Companies with poor tools will be left behind. Demand the best!!
"Can your internal AI tool beat ChatGPT/Claude on YOUR specific use case?"
Why Companies Keep Building Bad AI
Yes, data privacy matters. But most internal AI projects aren't handling anything more sensitive than what you'd put in an email. Microsoft, Google, and Anthropic have enterprise agreements that solve the privacy issue.
The Feature Factory Problem
Internal teams optimize for looking busy, not for results:
18-month roadmap for features that exist today commercially
Quarterly OKRs around "capabilities" instead of "user problems solved"
Demos designed to impress executives, not help users
"We're 80% there" (which means 6 months away, minimum)
Meanwhile, your competitors are shipping with commercial tools today.
Drive for internal tools and AI that drive results.
Weekly AI strategies for operating executives
— Brett
👉 Hit “Reply” and share your experience — I read every one!
Picture by xxx on Unsplash.