Goblins in prod, the messy middle of AI adoption, and everything is a harness now

**Ben Lloyd-Pearson** (0:05)
I don't know, Andrew, are you getting a Codex pet? I feel like it's only a matter of time until we have NFT-based identities for all of our bots.

**Andrew Ziegler** (0:13)
Oh, no.
That is not the direction I want this thing of all. I actually think it's funny to see things like Codex pets. For those who haven't seen it, it's a plugin for Codex that lets you add a terminal companion or a terminal friend as you're using the tool. We've obviously been seeing this before. A Cloud Code had buddy mode a few weeks ago or a month ago.

**Ben Lloyd-Pearson** (0:35)
I know you were lamenting the loss of yours. That was devastating. Yeah, I'm sure you're looking for a new pet.

**Andrew Ziegler** (0:43)
Trixels time, and this world was too short because the Cloud Code buddy feature only lasted about 10 sweet days. And so it's funny to then see the Codex folks over there and like, oh, you know, this is a grass is greener kind of moment where I'm like, oh, they're having fun over there with their pets. But you know what, I'm actually not really leaning into the whole trying to make a personality or trying to have it as this like thing that I talk with. A lot of times that my sessions are, you know, very ephemeral and they're anchored in durable sources of context and information. But like the sessions and agents themselves don't really glom much of an identity. Like even in the open-claw world, when that really hit the scene, there were parts of that that I stole and wanted to use from my own harness. And I certainly did, but, you know, the growing personality over time just wasn't really one of them.

**Ben Lloyd-Pearson** (1:33)
Well, that's not entirely true. Your agents do ask you to build them a profile picture at the very least. Well, that's seems like that is true.

**Andrew Ziegler** (1:44)
That is true. That is because all of my agents are represented as a little fish with a cowboy hat. For those that aren't in the Savvy Know, I post about them on LinkedIn because they certainly all have the same kind of shape and face. But actually, that is just because I have a really simple template that I go to NanaBanana and I make them a little fish with a hat. It's zero cognitive burden for me. I don't have to be like, oh, what's your name? What are you? Like, I'm not trying to put any kind of face on them, but it doesn't mean that I can't have fun with it. So, yeah, I'll go slap a fish with a cowboy hat on them.

**Ben Lloyd-Pearson** (2:15)
That's the kind of learnings you get here at the Friday Deploy, brought to you by LinearB. I'm your host, Ben Lloyd-Pearson.

**Andrew Ziegler** (2:22)
I'm your host, Andrew Ziegler.

**Ben Lloyd-Pearson** (2:24)
Yeah. This week, beyond Codex Pets, we have OpenAI's Goblin Invasion, organizational AI learning crisis, specs-maxing, and AI productivity myths. Tongue twisters aside, where are we going to start, Andrew?

**Andrew Ziegler** (2:40)
Let's talk about this story where goblins are overrunning our chats.
So this is a really fun post-mortem that came across our desk in the last week. If you've been on social media, you've probably seen fun prompts and outputs folks have shared from their usage of ChatGPT where it becomes obsessive in mentioning goblins, gremlins, and other creatures in its responses, rendering them in places where they don't belong, and goblin usage and goblin directed conversation spiking over 175 percent, which is an amazing thought to think that there's a dashboard or a metric inside of OpenAI that's tracking goblin. And it forced a lot of folks to add explicit anti-goblin instructions literally like a barricade on the town hall walls to keep goblins out of their outputs. So where did this come from? This is a really interesting dive into the reality of how LLMs are trained and where they get their performance from. So the problem on this originally originated from a personality training for a quote, nerdy ChatGPT preset where, you know, things like a creature references to a bestiary might come up every once in a while. It's the idea of like having a personality on clod or where like you'd probably be playing Dungeons and Dragons in a basement with it. So this output from this particular personality training ended up spreading to other models because it's pretty standard industry practice to use model outputs to train future models. So all it took was the existence of this nerdy ChatGPT preset somewhere deep in the training architecture of the models to slowly be pumping out goblin obsessed outputs and otherwise playing an infinite game of D&D in some virtual basement somewhere. And it really just dramatically polluted the output downstream for folks. So it really reminds us that this is a really big or a boros, right? Eating its own tail in terms of the kind of like performance we're getting. And it really calls to mind the importance of provenance and understanding the data that goes into your model.