Christian TonnyWriting

AI tools, Rwanda’s opportunity, and RTI

props to chatgpt 👏

While the giants duke it out, the opportunity gap widens.

The AI arms race hit new heights this week:

  • Google launched Gemini 2.5 with "thinking" capabilities that actually won on benchmarks
  • Google also released a Data Science agent in Colab that transforms how researchers analyze data.
  • OpenAI is reportedly pivoting from o3 to a unified GPT-5 approach and released their most advanced image generator yet (best I've ever tried, honestly)
  • Microsoft unleashed their Researcher and Analyst agents while also releasing MCP alongside OpenAI
  • Alibaba open-sourced Qwen2, making enterprise-grade AI deployment accessible to smaller players
  • Deepseek crashed through benchmarks with their powerful new V3 model
  • Reve appeared from nowhere and briefly topped leaderboards with their image generator—only for OpenAI to outpace them days later

In a word, there's no wall in AI—just relentless innovation accelerating by the day.

The Bubble Problem

Here's what fascinates me: Most African countries seem to operate in their own bubble when it comes to AI. Few are asking the critical question: "How can we compete in the AI arms race and deploy AI effectively for our unique challenges?" —despite having the most compelling use cases on earth.

This disconnect drives my upcoming article: “The Road to NST2: Rwanda’s AI-Powered Future.” Rwanda doesn’t need Silicon Valley’s playbook. We need our own human-centered AI strategy built for African challenges.

Why NST? Rwanda already has the foundation: - 97% internet coverage - 684 online government services - A digital-first policy framework

What I’m Building: RTI, Rwanda’s first AI-powered business intelligence platform, will deliver real-time market insights, regulatory navigation, and opportunity visualization. Front and back-end dashboards are progressing rapidly, and I’m seeking beta testers.

What I’m Learning: Re-reading “Steve Jobs” by walter Isaacson reminded me that true innovation happens at intersections. Jobs maintained unwavering focus on products that integrated hardware, software and services seamlessly. I want to built RTI in a way that will combine market data, regulatory knowledge, and AI—creating value impossible through any single component.

One Thing To Think About: The real AI advantage isn’t having the biggest models—it’s building the most relevant applications. Rwanda doesn’t need to compete on scale but on applying existing AI to our specific challenges.

Your turn: What African/Rwandan problems do you think AI is uniquely positioned to solve? Reply directly—I read every response.

Want early RTI access? Reply “RTI BETA” to join the priority list.

Read next