What’s the most important question right now?

How do we move from AI hype to AI literacy? I learn by building. AI literacy means flexing your critical-thinking muscle. At AJ+, we’re using multiple LLMs for research on our YouTube show “Historia vs Historia” because challenging colonial narratives requires diverse perspectives. Here’s what we discovered: DeepSeek surfaces Global South academics while ChatGPT defaults to Western white authors. Each model carries ideology.

That’s why I’m polyamorous with AI models. Critical thinking means understanding these tools aren’t neutral—they’re worldviews wrapped in algorithms.

What do you wish you had known a year ago?

Stop reading about AI. Start building with it.

Experimentation beats watching endless demos and reading articles. There’s a lot of marketing, you only realize what works when you try things yourself.

But let’s be real, this is a privilege. While I’m tinkering with new tools and design workflows, AJ+ teams cover Gaza’s ongoing genocide. Most of our journalists are crushed by the demands of a hungry algorithm and a fast-paced news cycle. The emotional toll and the vicarious trauma of being exposed to murdered children every day is very real.

AI experimentation requires upfront time investment (even if it is only on a personal level usage): data collection, prompt engineering… Not everyone has that luxury while running a newsroom.

What's capturing your attention?

AI’s potential to shape a new video grammar.

I’ve spent 10 years doing social video at AJ+—I’m an algorithm survivor watching platforms systematically deprioritize social justice journalism because it doesn’t sell ads.

Video is a language with distinct genres, like music. Social video has its own syntax—3-second opening hooks, on-screen text, 9:16 aspect ratios—completely different from documentary films or TV news. Think McLuhan: “the medium is the message.” With GenAI creating a new distribution medium that processes our content, visual journalism must “speak” differently. What are the first principles of video journalism in an AI world?

This feels like 2006 when I started experimenting with video on the internet. Back then, I thought interactivity was the first principle—that’s why I learned HTML, Flash, and PHP. It opened up interactive storytelling as a new genre to explore. It’s been amazing to see how digital video has evolved and changed (who knew back then that short vertical video and podcasts would rule our feeds?).

I still learn by making. Here’s what we’re building at AJ+ right now: a Premiere Pro plugin that reduces multilingual adaptation from hours to minutes while keeping editors in control. We’re also experimenting with AI motion graphics for breaking news and in-camera visual effects using generative AI during production—capturing immersive visuals in real-time rather than fixing them in post. I am watching AI-powered animation with curiosity because it stretches the language of video. The question I keep asking is simple: how will visual journalism “speak” inside AI-mediated environments?