Mario Thomas
@mariothomas
Followers
4K
Following
13K
Media
380
Statuses
3K
Head of Applied AI & Emerging Tech Strategy @AWSCloud | Chartered Director & Fellow @The_IoD | AI, Cloud, & Data Leader | @LSEnews Alumnus | Opinions my own
Llantwit Major, Wales
Joined February 2009
🧠🚀 This is exactly what AI enablement looks like in practice. A 17-year-old, six weeks, and a well-designed ML model - outpacing decades of traditional analysis. The tools are here. The barriers to scientific discovery are falling. The question for every organisation: are you
🚨 A student in the US just discovered MILLIONS of new space objects. The astronomy world was recently shaken by a discovery from an unexpected source: a teenager still in high school. Matteo Paz, a student from Pasadena, utilized archival data from NASA’s retired NEOWISE
0
0
0
The behind-the-meter transition I wrote about two weeks ago — now in the FT. The supply story is clear. The strategic question: what happens when surplus capacity turns energy consumers into grid actors? 👉
Data centres turn to aircraft engines to avoid grid connection delays
0
0
1
This mirrors what I've been hearing across industries. The question "Do you actually need generative AI?" isn't contrarian anymore, it's becoming standard due diligence. More on what this means for Boards in my next article.
Exclusive: Salesforce says trust in generative AI has declined. Read more from @AaronpHolmes and @KevKubernetes 👇 https://t.co/Y6PnjnipF4
0
0
0
📊 America's AI ambitions face a 19GW power shortfall by 2028 — over 40% of what data centres need. The @FT's new analysis on America's AI power crunch confirms what I've been exploring in my UK sovereignty series: energy access is becoming the competitive moat that determines
0
0
0
Beyond the buzz, agentic AI demands conscious delegation. Pilot wisely, govern minimally but lovably, measure beyond tech demos. Boards treating this as procurement? Recipe for that 40% failure rate. See how to do it here:
0
0
0
Strategic choice time: agentic AI adds value where scale/speed beat expertise—like routine queries or doc analysis. Keep humans for proprietary knowledge or high-stakes calls. Multi-speed adoption means mixing both. How's that balance working for you?
0
0
1
The legitimate innovation in agentic AI? Not the models - the architectures for safe, scalable delegation. Autonomous planning, tool execution, persistent memory. But Boards must ask: does this compound advantage, or just complicate? Your experiences?
0
0
0
Stage 3: human-in-the-loop. AI iterates, but you review and tweak each round. Solid for complex stuff. Agentic (stage 4) cuts you out entirely for the right tasks. Boards, which functions in your org scream for this handover?
0
0
0
In stage 2, AI's your on-demand copilot—quick ideas or drafts, but humans hold all judgment. Handy for starters, but agentic takes it further. Boards: is this where most of your AI sits today? Or are you pushing to loops?
0
0
0
Stage 1 of agentic evolution: all human work, no AI. Safe, but slow. Boards clinging here while competitors loop ahead? Time to assess where automation adds speed without losing control. What's holding your org back?
0
0
0