What's next for Sources

Publishing four days a week, a live interview series, and raising subscription prices.

Jan 06, 2026

It’s Monday, Jan 5th. I’ve landed in Las Vegas for CES, where I plan to spend as little time on the showfloor as possible. I’m here for a bunch of meetings, a podcast recording, and to hopefully pass on what attendees are actually buzzing about all week.

First, 3 things you should know:

Yes, we are already five days into 2026.
Smart rings are hot at CES this year: Sandbar, Pebble, Oura, and others are here promoting their latest hardware.
The 1990s are alive and well this week in Las Vegas. Snap is throwing a party with Snoop Dogg tonight, and Public Enemy is performing at the Yahoo party later this week.

Mentioned in this issue: Fidji Simo, Jensen Huang, Yann Lecun, Adam Mosseri, Jerry Tworek, Casey Newton, and Dan Wang.

Sources 2.0

In mid-September, I left my full-time media job to launch this newsletter and the ACCESS podcast.

Since then, Sources has drawn thousands of subscribers from the most influential companies in tech, finance, and media. CNBC, TechCrunch, TPBN, The Information, Variety, Fast Company, and other outlets have covered my scoops. I’ve published exclusive interviews with tech leaders like Demis Hassabis, Greg Brockman, and Daniel Ek, and I've run my first brand partnership with Brex.

Now, it’s time for Sources 2.0: I’m going to start publishing four times a week, up from the two to three issues per week I’ve been publishing over the past few months. I’m also increasing the subscription price to $99 a year or $10 a month. Existing paid subscribers, most of whom are on an annual plan, will remain at the original rate unless they cancel and resubscribe.

Two issues of Sources per week will remain free to read for up to 7 days after publication. I intended to always put my best stuff behind the paywall, but I still want Sources to be a valuable product for the vast majority of subscribers who are on the free tier. Publishing days will be Monday, Tuesday, Wednesday, and Thursday in the evening Pacific time, excluding holidays and vacation time I’ll share in advance.

After working in various newsrooms, I see an opportunity to build a new publication focused on the AI race, informed by my decade-plus of covering tech. My goal is not necessarily to be the only place you get AI news, but to bring relevant, original reporting and unique perspectives you can’t find anywhere else. My current plan is for each issue to roughly follow the format of this one: an intro, a lead story, a few shorter items, a rotating section depending on the day, and some more links.

I’ve got a lot of plans in store for 2026 beyond the written newsletter, too. After a successful launch party in New York City, I’m working on launch events in San Francisco and Los Angeles. I’ll also be in Davos, Switzerland, later this month at the World Economic Forum for the inaugural series of Sources Live interviews. More to come on that front soon.

Despite my best efforts to prepare, I had no idea what going independent would feel like when I took the plunge last year, or if this newsletter would show early signs of life. I still have a lot to learn about building a lasting media property, but I do know this: both original reporting and showing up consistently have worked so far. So that’s what I’m doubling down on.

Elsewhere:

OpenAI is going to start testing ads in ChatGPT: That news was, for some reason, left out of the public product roadmap that CEO of Applications Fidji Simo shared today, but sources say she mentioned that the company planned to start testing ads in an internal version shared with employees. Her public post is still very much worth reading, as it’s the first broad articulation of what she’s focusing on at OpenAI. I expect that “capability overhang” will become much more of a buzzword as the year progresses. (On the ads piece: Over the holiday break, OpenAI told The Information that it planned to “evaluate whether a conversation has commercial intent and then pull up the most relevant ads in ChatGPT responses.” Something has to pay for all those GPUs!)

Nvidia kicks off CES: As a longtime attendee, there was something strange about watching Nvidia’s kickoff keynote at CES today begin with a chip analyst talking about the AI bubble. The consumer hardware part of CES fizzled a long time ago, so I suppose it’s only natural for AI to fill the void of compelling gadgets. CEO Jensen Huang broke out his shinier jacket for his annual keynote, which mostly served as a general state of the union for AI. The main news he shared was that Vera Rubin is in “full production,” which isn’t surprising if you’ve been following the company’s previous teases. Nvidia is also launching its own AI model for autonomous vehicles, which Huang said will be released with Mercedes in the US this year. One of his biggest customers has already taken notice:
Elon Musk@elonmusk
@SawyerMerritt Well that’s just exactly what Tesla is doing 😂 What they will find is that it’s easy to get to 99% and then super hard to solve the long tail of the distribution.
12:01 AM · Jan 6, 2026 · 265K Views
455 Replies · 395 Reposts · 7.79K Likes

Overheard

Yann Lecun throwing bombs at Meta on his way out the door to start Advanced Machine Intelligence:

“We had a lot of new ideas and really cool stuff that they should implement. But they were just going for things that were essentially safe and proved. When you do this, you fall behind… A lot of people have left, a lot of people who haven’t yet left will leave… You don’t tell a researcher what to do. You certainly don’t tell a researcher like me what to do.”

Instagram chief Adam Mosseri wrote a lengthy post about the future of social media and AI that I can’t stop thinking about:

Authenticity is fast becoming a scarce resource, which will in turn drive more demand for creator content, not less. The creators who succeed will be those who figure out how to maintain their authenticity whether or not they adopt new technologies. That’s harder now—not easier—because everyone can simulate authenticity. The bar is going to shift from “can you create?” to “can you make something that only you could create?” That’s the new gate.

OpenAI research lead Jerry Tworek’s farewell note to colleagues (emphasis added):

It’s been a blast, and while I am leaving to try and explore types of research that are hard to do at OpenAl, this is a special company and a special place in the world that already has a forever place in history of humankind.

Discussion about this post

Ready for more?