FOIL Disclosure #2 | Three AI shifts worth your attention
This is a copy of FOIL Disclosure, a monthly newsletter written by Mish Naik, Head of Strategy. To get notified of the next one a month in advance, subscribe to the FOIL Substack here.
We've been sifting through the noise to find those gems that make you pause and think, "Wait, that's actually quite interesting." Perhaps sparking your curiosity and some meaningful conversations.
Grab yourself something warm (or refreshingly cold) and let’s explore what’s caught our eye this month.
1. Agentic AI: From Assistant to Collaborator
The intrigue: Remember when AI just sat there waiting for instructions? That era is fading fast. Agentic AI is beginning to show initiative—making decisions, taking actions, and even completing complex tasks with minimal human oversight. It's like watching a colleague gradually take ownership of their role.
So what is it? Harvard Business Review has written a great piece unpacking how agentic AI is quietly reshaping our relationship with technology in the workplace. [Harvard Business Review]
Real-world ripples: Anthropic's experiments with virtual employees offer a fascinating glimpse into what happens when AI systems start operating with greater autonomy. Worth a read if you're wondering what your team could look like in the not-so-distant future. [Axios]
Lessons from Carnegie Mellon: Carnegie Mellon's recent "TheAgentCompany" simulation reveals something rather fascinating about AI's workplace readiness. These digital colleagues can autonomously handle about 24% of workplace tasks. But here's the curious part: they shine with software engineering challenges yet stumble through administrative tasks requiring complex UI navigation or social interactions. They're still finding their footing with complex, multi-step challenges. Makes you wonder which tasks in your organisation might be first to transform? [Carnegie Mellon]
Points to ponder:
Where in your organisation could you benefit from a collaborator that never sleeps?
What appetite would you have for to decision-making responsibility to shift to systems enabling them to start making judgment calls?
2. The Human Side of AI Adoption: It's Complicated
The intrigue: The technical implementation of AI often proves simpler than navigating the human response to it. The psychology behind AI adoption is becoming as crucial as the technology itself.
The widening divide: Azeem Azhar presents a compelling case that AI might actually be amplifying differences between top performers and everyone else, challenging our assumption that AI democratizes capabilities. A thought-provoking read that might shift how you approach your AI strategy. [Project Syndicate]
Bold moves in hiring: Shopify's CEO Tobi Lütke has flipped traditional hiring on its head: teams must now prove AI can't do the job before bringing on new talent. Radical? Perhaps. Worth considering? Absolutely. [Forbes]
Points to ponder:
What would adopting a "prove AI can't do it first" approach like Shopify's reveal about your current workforce allocation?
How might you balance the efficiency gains from AI with the potential impacts on team dynamics and employee sentiment?
3. AI in Education: Rewiring Future Talent
The intrigue: Educational Institutions and Governments are grappling with AI's implications in real-time. Their approaches today will shape the talent pipeline you'll be drawing from tomorrow.
The adoption divide: Anthropic's revealing report shows computer science students embracing AI tools enthusiastically while students from other courses are lagging. What might this mean for interdisciplinary collaboration in future workplaces? [Anthropic]
China’s ambitions: China's bold move to make AI education mandatory for all schoolchildren from September 2025 raises fascinating questions about global competitiveness and different approaches to cultivating AI literacy. [Engineerine]
US Policy Shift: President Trump signed an executive order that establishes a White House task force that promotes public-private partnerships, and introduces initiatives like a "Presidential AI Challenge" to foster AI literacy among students. [Forbes]
Cognitive landscapes: "I prompt, therefore I am" explores how our thinking processes are being reshaped by AI collaboration—particularly in learning environments. A genuinely thought-provoking read on what it means to think in the age of AI. [Swiftscale]
Points to ponder
How might organisations benefit from (or struggle with) graduates who've developed fundamentally different cognitive approaches?
Which human capabilities become more valuable when AI handles predictable cognitive tasks?
Is there a balance to be struck between AI proficiency and preserving independent problem-solving abilities?
No doubt this evokes some interesting thinking amongst your peers, like it did for us. Drop a comment to share your thoughts or even better, book a conversation with us. Don’t forget to subscribe for more of this next month.
Until our next curious wander through the landscape,
Mish and the FOIL team
P.S. What's captured your attention lately? Drop us a note—we're always keen to explore new horizons together.