In this issue: Three years since ChatGPT launched, and we’re caught between early days and exhaustion. Why AI feels like an imposition. What quality journalism needs to survive the chatbot era. Sergei Yakupov on why “why” still matters more than “how.” Plus: Gemini turns cookbooks into shopping lists, and merch for people who really shouldn’t be driving Cybertrucks.
What we’re talking about: It’s been three years since ChatGPT. On one hand, we’re still in the early stages. On the other, fatigue is creeping in: another AI conference? Another panel? Another post on LinkedIn?
If you spend your days wrangling prompts or babysitting agentic systems, the next consultant showing up with a “disruption deck” feels almost adorable. Yes, the tech is wild. Yes, it’s powerful. But the hype machine keeps selling us instant revolution while the real world is like, “let’s circle back in Q1.”
At a recent event, Isabel Lerch asked me to complete the sentence “Artificial Intelligence is…” My answer: an imposition. The industry foists faulty chatbots on us. I can’t stop thinking about the posters that one tech company used to print and distribute with such pride: “Move fast and break things.”
The question then: what do we do with AI? Which means training teams, buying tools, building systems. None of that is cheap. Even more important: How do we position ourselves in this new landscape? For quality journalism, it means being different, better, more useful than a hyper-personalized chatbot.
For many outlets, the timing is brutal: a century of printing news, barely adjusted to digital, and now another disruption hits. But I actually think journalism might come out stronger. With a clearer sense of mission. Real audiences stick around when you’re honest with them, when you give them something a bot can’t. Maybe that’s the whole point.
JournalismAI Festival 2025: No more FOMO – 23 recorded sessions are now available. Jane Barrett on what’s coming next? Chris Moran on Newsroom Singularity? Sorry, can’t right now, gonna be watching for a while.
The Thinking Game: This documentary takes us behind the scenes with Google DeepMind chief Demis Hassabis, showing how they taught AI to play video games, chess, and Go, before moving on to protein folding. It’s compellingly told, though it sidesteps what was happening simultaneously at OpenAI. Free on YouTube.
OpenAI co-founder Ilya Sutskever discusses the current state of generative AI, and talks about his new company SSI and what he plans to do with the three billion dollars he’s raised: develop learning AI systems. (Dwarkesh Podcast)
To mark ChatGPT’s third birthday, the Guardian’s Robert Booth tours Silicon Valley in search of the future: “Everyone is working all the time. Even the folks who are very wealthy now … all they do is work,” says someone from DeepMind.
In media outlets like n+1, resistance against AI is taking shape: “When we use generative AI, we consent to the appropriation of our intellectual property by data scrapers. We stuff the pockets of oligarchs with even more money. (…) There’s still time to disenchant AI, provincialize it, make it uncompelling and uncool.”
Slop Evader: A search tool that will only return content created before ChatGPT’s first public release on November 30, 2022. (Tega Brain)
And now: During a trip to Riga for the Baltic Centre for Media Excellence, I finally got to meet the person behind AI For Journalism face-to-face. Sergei Yakupov maintains an ever-expanding collection of use cases and industry news.
The question is “Why.” Tech can answer almost any “How” question, but we still have to ask ourselves the “Why” questions. Why do we need to do this? Why will the audience stay with us? Why do we need this platform, this tool, these people?
“Why” is still the most important question in journalism, in media, and in product management. Without it all “how” questions don’t mean much.
A year from now, what will we be shaking our heads at?
I actually have at least two answers to this question. We’re living in pure tech chaos (thanks, AI disruption). But I think that over the next 12 months the ecosystem will start to clear up a little. A year from now it might simply be easier to navigate because the number of choices will shrink: big tech will get even bigger, offer even more, and most of us won’t need to look far beyond them.
But I also have a second answer (more product focused probably). In a year news media might finally be able to choose between relying on big tech or using open-source tools with comparable power (I believe that the capabilities of OpenAI-type models and open-source models will become increasingly comparable). We’ve already talked about the problems around AI infrastructure, and I hope that once the pace of AI development slows down (because of compute limits or worries about stumbling into AGI) we’ll hit the next phase: AI infrastructure becoming simpler, cheaper, and more diverse that helps us to look closely to open source LLMs.
If that happens, newsrooms will stop asking how to implement AI and will start asking why and where it actually makes sense. So AI will become more regular thing for news media. Instead of constantly trying to stay “up-to-date” and hopping from one tool to another, we’ll be able to focus on building real products and making intentional choices. I don’t know if this is exactly what will happen, but I’m 100% sure it’s the direction we need to be moving toward right now.
What's a good hobby to pick up?
I live near the ocean, so I often walk along the shore and one day I started picking up old pieces of driftwood because I thought: “Hey, this looks like a whale,” or “This could be part of an old rusty lighthouse.” Eventually those random pieces turned into small driftwood sculptures — rusty lighthouses on rocks, little surfer villages, sea creatures, and so on.
This hobby helps me to go beyond the digital world I spend most of my time in. It gives me something tangible to work with, and it lets me extract images out of my head and turn them into something real. And frankly I do love the rusty aesthetic of old lighthouses.
Hands on: There’s so much to report, about vibecoding with Replit, about the attempt to push WordPress entries into a vector database. But first: cookbooks. Because I just wanted something that worked without overthinking it.
I love cookbooks! Though some use these peculiar measurements: ounces, quarts, cups. (Americans likely feel the same way about grams and milliliters.) The other day, in a rush, I took a quick photo of a cookbook page and asked Gemini 3 to convert the measurements and generate a shopping list. Gemini not only did the conversion, it automatically created a Google Notes entry.
Give me a shopping list, convert the measurements to metric / US customary system
For anyone who still needs inspiration for the holidays:
What is a good vegetarian substitute?
It works. And it doesn’t spoil the joy of cooking.
December protocol: My website got its own little Advent upgrade. Every day there’s a new surprise waiting, which feels very “internet 2012” in the best way.
One more thing: This latest drop from some streetwear brand popped up in my feed. I can’t really tell who it’s actually for. Cybertruck drivers? Crypto guys? Students praying their proctor doesn’t notice the very obviously-not-official hoodie they’re wearing during exams?
It feels like merch for a subculture that maybe exists, or maybe we all just hallucinated on X at 2 a.m. Is any of it real? Or is it just another case of a brand trying to manifest a vibe into existence? Either way, it’s definitely not official merch.
This is THEFUTURE.
Subscribe to THEFUTURE
Get this newsletter in your inbox. For free.
You want to join THEFUTURE. Please check your mail for confirmation.
