Platformer Reinvents Its Model in Response to the Rise of AI

Reflections on the Evolution of Platformer

While checking my messages on Signal, I received an interesting tip for an article. However, instead of the usual excitement, I felt apprehension. When would I be able to make the necessary calls to verify this information? Would I have time to meet sources in person? Since launching my daily newsletter in 2017, I have often felt this pressure. My approach has always been to report scoops when possible, supplemented by news analysis. This has led to an imbalance, favoring analysis over in-depth investigation.

Since the 2016 U.S. presidential election and the backlash against tech companies, I found a niche in publishing link roundups on technology and democracy. However, the newsletter landscape has become saturated. Many writers have left newsrooms to write for their own audiences. Newsletters, once ancillary, are now central for many publishers. Yet, the decline of the web and the struggles of large media companies have reduced the available tech journalism.

The rise of artificial intelligence this year has introduced systems that encroach on our work. In January, I experimented with an automated daily briefing of link summaries. It works as well as I do at finding interesting stories, and it happens automatically while I sleep. Link aggregation, while never the most valuable work, has seen its worth diminish. More and more people are using personal agents to write news summaries. An increasing number of Substack publications seem to be generated by AI.

I expect that some, particularly leaders and policy professionals, will begin to rely on AI for news analysis. Chatbots are becoming more accurate in their responses regarding the implications of news. For now, they have less authority than experts who write paid newsletters. But having been a reporter since 2002, I have seen how the internet dequalifies and replaces journalists. I can imagine a future where morning and afternoon newsletters are replaced by personalized briefings from agents.

Necessary Adaptation for Platformer

If the three pillars of Platformer are original reporting, news analysis, and link aggregation, it seems that one of these pillars is already commodified and another is on its way to being so. We have tried to adapt to these changes, but I believe we need to act more quickly. In September, I expressed my desire to spend more time on reporting. Subscribers supported this change, reminding me that quality trumps quantity.

I am proud of the original reporting we have done: unreported internal conversations about Meta's AI spyware, a potential disengagement from Meta's Oversight Board, and the closure of OpenAI's mission alignment team. We have also generated interest in our firsthand experiences with AI.

We have done our best with our current schedule. But occasional days off are not enough to give me what I truly want — the flexibility to leave my office for several hours, to develop stories over several days, and to spend a slow news day reporting rather than gathering a column.

Today, we are starting an experiment to see what this version of Platformer would look like. Free subscribers will still receive one column per week. Paid subscribers will get an additional column on Thursdays, which we consider a reporter's notebook: what I’m hearing, what we’re working on, a preview of Hard Fork, and a mailbox. Some may resemble traditional columns; others may be bolder.

Paid subscribers will also receive additional stories and analyses as we write them. This is the biggest change: instead of promising to publish on a fixed schedule, we will publish when we discover something interesting or when we want to help you understand the bigger story of the day.

In practice, I think there will be many weeks where paid subscribers will still hear from us three times a week. But for all the reasons mentioned, we need to change Platformer so that our schedule serves journalism. For too long, it has been the opposite.

Adjustments and Future Plans

We are making a few other adjustments. Side Quests, our end-of-column section with all the interesting links, is going away. Techmeme does this job better than we do, 24/7. In the future, we want to reserve our energy for advancing the story.

The Following section, launched in September to inform about the day's news with our thoughts and comments from influential people, remains. Readers appreciate it greatly, and we will continue to send Following editions to paid subscribers to help understand the big stories on days when we are working on other topics. You will continue to find them after our columns.

Finally, we are working on a major project: a limited series of conversations about the future of AI and work, which will launch next month in text, audio, and video formats. The goal is to capture the current uncertainty and the latent conflict between workers and managers, as covered earlier this year in the AI-related productivity paradox. We have exciting discussions planned, and I look forward to sharing more with you soon. In the meantime, it's a good time to add the Platformer feed to your favorite podcast player.

As always, I would love to hear your thoughts on these changes. I am sincere in calling this an experiment — we plan to iterate on this new approach over the coming months and can always change or revert based on reader feedback. Honestly, I find these changes somewhat terrifying, as they replace the logic of a three-times-a-week column production line with something wilder and less certain.

But as a brilliant Silicon Valley CEO once said: only the paranoid survive. In a world where everyone has an opinion but almost no one has a second source, we are betting that the value of tech journalism is shifting from aggregation and predictability to original reporting and surprise.

Thank you to everyone who has supported Platformer so far. And for all the others, if this new chapter sounds interesting to you, we would be thrilled to have you join us as well. In the meantime, I have a tip to follow.