Paying Attention #6: Meta's Mythos, Claude Tag, Fable's Return + GPT-5.6 Launch

Happy Sunday everyone!

Hot Atlanta Summer has arrived. It’s been in the low 90s all weekend, or the low to mid 30s for my international audience. My wife and I will see England play the DR Congo in the World Cup on Wednesday. A true bucket list experience and we can’t wait.

Now let’s get into three things I paid attention to this week.

1. Meta’s Mythos

Dan McAteer@daniel_mac8

I got a big, fat, juicy AI rumor from my friend. Apparently Meta trained a Mythos-level model. And Zuck and team are so back. No timeline yet but probably Q3/Q4 release.

2:11 PM · Jun 23, 2026 · 203K Views

---

58 Replies · 18 Reposts · 288 Likes

What happened

A friend of mine in the know mentioned that Meta is almost done training a Mythos level model.

Why it held my attention

As we’ve seen over the last few weeks, it’s important to have a diversity of labs capable of training frontier AI models. Meta is also the only lab that hasn’t agreed to the Trump admin’s “voluntary” model testing.

What I’m carrying forward

The question is, how fast can Meta ship a SoTA model? Because Mythos of today won’t be the same thing as Mythos of even three months from now.

2. Claude Tag

Dan McAteer@daniel_mac8

Claude tag is live in our company Slack. It's fantastic. It does feel like a new paradigm. I've heard Anthropic employees say that you should treat Claude like a coworker. Claude Tag makes that easy. Pro tip from @trq212: create a personal private channel for Claude Tag.

Claude @claudeai

Introducing Claude Tag, a new way for teams to work with Claude. In Slack, Claude joins as a team member with access to the channels and tools you choose. Tag Claude in and delegate tasks to it while you focus on other work.

9:02 PM · Jun 23, 2026 · 69.6K Views

---

26 Replies · 7 Reposts · 170 Likes

What happened

Anthropic released “Claude Tag”. It is a proactive, stateful, flexible instance of Claude running in your company Slack. Andrej Karpathy called it an “org-level harness”.

Why it held my attention

There was a negative consensus on X for Claude Tag. I don’t get it. We implemented it in our company Slack and I was getting value out of it from day one. That was just the first few days of using it. I do understand that it’s metered usage, which means you need to be responsible with how you use it. My rule of thumb is: does this task have a realistic chance of creating value for the company?

What I’m carrying forward

This is the first implementation of AI agents that I’m aware of that feels like an attempt at a “digital employee”. There’s a strong possibility that we need to get used to this way of working.

3. Fable’s Return and GPT-5.6 Launch

Dan McAteer@daniel_mac8

My prediction: GPT-5.6 ships and Fable returns this week. Axios reports the government will lift export controls on Fable as soon as this week, and they're unlikely to favor Anthropic over OpenAI. Fable is the most capable model we have. It's also slow and expensive. GPT-5.6

Mikhail Parakhin @MParakhin

Not as relevant now :-(: I had an opportunity to deeply test both Fable 5 and GPT-5.6 Max. 5.6 is clearly better than Opus 4.8 at everything (slightly faster, too, though that depends on the load). Vis-a-vie Fable, it is clearly worse on coding, but better on agentic workloads. I

11:59 AM · Jun 28, 2026 · 42.2K Views

---

26 Replies · 14 Reposts · 294 Likes

What happened

Axios reported Saturday that Anthropic is in talks with the US, and that Fable could return as early as next week. Mythos was okayed for select partner organizations at the end of last week. Also, GPT-5.6 was announced, but not released on Friday.

Why it held my attention

There’s a good chance that Fable will return this week, and GPT-5.6 will be released along with Fable’s return. In fact, that’s my prediction. I don’t think the Trump admin wants to come off as capricious and playing favorites. Even if, in reality, they are, they don’t want to be seen that way.

What I’m carrying forward

This whole saga of the US government vibe-governing the Singularity is unsustainable. We need legitimate, law-governed regulation for AI that is established by lawmakers and has clear stipulations.