Ep 820: The Most Important AI Model You’ll Probably Never Use That Just Dropped

You've probably never heard of Inkling. 

It's the newest (and first) model from Thinking Machines Labs, and it could very well be a small snowball that picks up major momentum in today's enterprise AI landscape. 

If you haven’t heard of Thinking Machines, they’re led by Mira Murati, the former CTO at OpenAI. 

The big bet with Inkling? 

The future of AI could be using smaller models fine-tuned and optimized for smaller tasks. 

Will it work? 

Tune in live as we dive in. 

The Most Important AI Model You’ll Probably Never Use That Just Dropped -- An Everyday AI Chat With Jordan Wilson

Newsletter: Sign up for our free daily newsletter
More on this Episode: Episode Page
Today's Episode on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.

Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Website: YourEverydayAI.com
Email The Show: info@youreverydayai.com
Connect with Jordan on LinkedIn

Topics Covered in This Episode:

1. Inkling AI Model Launch Overview
2. Thinking Machines Lab Leadership Highlight
3. Inkling's Multimodal and Agentic Capabilities
4. Open Source vs. Proprietary AI Models
5. Enterprise Procurement with American AI Models
6. AI Fine Tuning as a Service (Tinker)
7. Benchmark Scores: Inkling vs. Frontier Models
8. Customization and Model Shopping for Enterprises
9. AI Token Costs Driving Model Efficiency
10. Bridgewater Case Study: AI Model Customization
11. Frontier Models Enabling Efficient Fine-Tuning
12. Future Trends: Specialized Small Language Models

Timestamps:

00:00 Inkling: A new AI model release

05:43 Inkling AI model details

09:08 China's dominance in open source AI

11:48 Launch and model updates discussed

15:21 Concerns over using Chinese open-source models

19:06 Training smaller AI models

20:22 Using GPT for AI Model Training

23:54 Predicting Rise of Small Language Models

28:38 Choosing the right AI model

Keywords: 

Inkling, Thinking Machines Lab, Meera Muradi, former OpenAI CTO, open source AI model, American AI model, fine tuning as a service, enterprise AI, multimodal AI, agentic models, customizable AI, Tinker, enterprise distribution, model procurement, Chinese open source models, strategic reset, model overhang, capabilities gap, AI model shopping, model routing, cost-conscious enterprises, artificial intelligence index, 975 billion parameter model, text-image-audio AI, open weights, proprietary AI models, customization accessibility, small language models, AI workflows, context window, Bridgewater use case, model distillation, GPU infrastructure, API costs, token efficiency, fine-tuned models, post training, AI competitive leverage, recurring financial judgment, AI benchmarks, middle tier models, automated model evaluation, privacy and workflow mapping, economical AI models, model rental, model routing automation.

Ready for ROI on GenAI? Go to youreverydayai.com/partner 

Transcript

SPEAKER_00 0:17

One of the more important AI models you've probably never heard of and likely won't use just got released. This may get slept on, but Inkling is a huge release from Thinking Machines Lab. So if you haven't heard of Thinking Machines, they're led by Mira Marathi, the former CTO of OpenAI. So why is Inkling maybe the most important AI model you probably won't use? Because it's now the best open source model from an American company, and Thinking Machines is betting on the future of AI, fine-tuning as a service. The model itself, though, it's multimodal, it's agentic, it's customizable, and it's available through enterprise distribution on day one. Inkling doesn't need to beat every Chinese model or Claude Fable to be relevant. It only needs to pique the interest of a few cost-conscious enterprises to not only be extremely profitable, but to also help shift the conversation around enterprise AI. Frontier models are becoming so powerful they can actually fine-tune smaller, practical models without much iteration or without even much expertise. So, has a new category re-emerged? Is fine-tuning back? And what might think machines' first major product mean for the broader AI competition? Well, here's the big picture. Inkling could change how enterprises buy AI. So this was just released Wednesday, so hours ago. And it is not the smartest model, but it is strategically important. I think it gives companies a credible American open weight model alternative to Chinese models. And inkling could be the best general purpose open model because it is multimodal, right? A lot of the open source, open weight Chinese models aren't. Many of them are text only. So Inkling, it is multimodal, it is a gentic. And as workflows in the real world lag so far behind model capabilities, I do think that there's a real market for bespoke middle-of-the-pack AI. So on today's show, here's what you're gonna learn. You're gonna learn why an open model matters, even if you're never gonna deploy it. You're gonna understand how Inkling reopens enterprise AI options beyond just Chinese open models. You're gonna know why frontier models could make fine-tuning practical beyond research teams. And you're gonna know how model shopping is gonna change budgets, vendors, and competitive leverage. All right, let's get to it. Welcome to Everyday AI. If you're new here, my name is Jordan Wilson. We do this every day. It's your daily live stream podcast and free daily newsletter helping business leaders like you and me keep up with the nonstop AI updates because my gosh, can't take an hour off. I help you decide what's important. I tell you how to use all this information to grow your company and your career. So it starts here with the unedited, unscripted live stream podcast. But make sure you go to our website at your everydayai.com. That's your cheat code. We're gonna be not just recapping the highlights from today's show, but go sign up for our free daily newsletter. We're gonna be re uh giving you everything else that you need to know that's happening in the world of AI today. Because yeah, it's one of those things you gotta like go and work it out. It's a muscle. Use it every single day. Uh, and yeah, all the AI news will be in our newsletter. All right, let's get into it. Live stream audience, good to see you. Uh, Adam joining us from St. St. Louis, Jose from Santiago, uh, Angie joining from Montana, Amiko, Tokyo, Brian, what's up, Brian? Joining from Minnesota. Uh, so uh, let's talk about maybe the most important AI model you'll probably never use. So there's a lot of different factors that have been compounding over the last, I would say, three months. And to put it in a very short uh summary, frontier models are probably for the most part too much for many enterprises, right? I'll say this I, you know, probably the Fortune 500, they can squeeze as much juice and uh, you know, make it worth the cost. But I'd say for many companies, especially those probably in the Fortune 5 or like the Fortune 501 to the 5000, right? So this isn't everyone. Um, you know, I don't think this changes the equation for every single company, every single business out there, because so many people are still gonna want, you know, their Chat GPT Enterprise, their, you know, Gemini, their Claude, their co-pilot. They want to make things easy, and they're not necessarily, you know, looking at, okay, how do open models or fine-tuning models change our strategy? But for so many in our audience, it does. And this is a really big deal. So here's a little bit more about the model itself. So, model, uh, the model is from Thinking Machines and it's called Inkling. It is a 975 billion parameter model, and it's multimodal. So, text, images, and audio. Uh, what is interesting as well, uh, previously Thinking Machines did demo kind of a similar dual purpose or uh you know two-way street um audio model as well, so that can kind of hear and listen as well as talk at the same time, like OpenAI's new GPT live. So Inkling is led by former OpenAI CTO Miro Marathi, and its quick arrival raises, I think, the next big question, which is how good is it? Well, here's from the company themselves from their release. They say our model called Inkling is a mixture of experts transformer with 975 billion total parameters, 41 billion active. It supports a context window of up to 1 million tokens. It was pre-trained on 45 trillion tokens of text, images, audio, and video. It is the first in a family of models of different sizes. Uh, alongside it, we are sharing a preview of Inkling Small, a lighter weight model with 12 billion active parameters, trained with a smaller recipe that achieves strong performance with even lower costs and latency. Inkling reasons natively over text, images, and audio and balances costs with performance through efficient and controllable thinking effort. We trained it to be a broad balanced foundation model, strong across many domains, flexible enough to adapt. Inkling is not the strongest over model overall model available today, open or closed. Instead, a combination of qualities makes it a good open weights base for customization, multimodal capabilities, efficient thinking, and availability on Tinker for fine tuning. Inkling is just the start, our first release in a model family. We will continue to build on. We want to make customization accessible for more use cases. So Inkling is available for fine tuning on Tinker today. Uh, picking the right base model to fine-tune is a qualitative judgment that combines measurable benchmarks with a unique feel of a model that comes from playing with it. To enable the latter, we're adding the Inkling Playground in the Tinker Council, a developer-facing interface for chatting with Inkling. To show what customization means in practice, we asked Inkling to fine-tune itself using Tinker. The model wrote its own fine-tuning job, ran it, and evaluated the result. So, long story, kind of short, right? If you want to go use Tinkling, uh Tinkling, right? And I don't, I don't know why. The combination of Thinking Machines Lab and Tinker, it just doesn't roll off the ton, uh, right. But if you want to go try it out, so they do have like a dev council where you can go quote unquote chat, which is interesting that they just didn't release a chat version. Um, anyways, this is, I think, a really big deal. And here's one of the reasons why. Uh, yes, benchmarks. So, podcast audience is showing the artificial analysis intelligence index. Uh, but this is uh shaded here by our uh closed proprietary and our open weights models. So let me zoom way out for our very non-technical audience, or if you're very new to AI. What's the difference? What's proprietary open source, right? Proprietary are models that you can't really modify them to the core. You can add custom instructions to them and you know change their behavior that way, but you can't really change the foundations of these proprietary models, right? Those are the uh the Clauds, the GPTs, the Geminis, etc., right? Then you have these open source or open weight models. And these are ones uh that for the most part China has been absolutely dominating on. And they've been doing it through distillation, which is uh, you know, not exactly uh uh something the American labs are happy with. But the you know, the Chinese labs are essentially, you know, stealing or borrowing whatever you might call it, uh, the work of the uh American labs and making their own versions of these models. And then they serve those, right? So you can if if you're a big company and if you have the the server racks, you can use these models and the open uh open source, open weight models. If you have the infrastructure, you can download them and run them 24-7 and not really pay any additional costs if you have that capacity. So that is the um the allure of these open source models. Or as consumer hardware becomes more capable, being able to run some of these locally, all right. So some of these models, not necessarily the ones I'm showing here uh on screen, but you know, some of these open source models, if you do have a very expensive, very beefy uh, you know, machine, you can run a slower version of them. So that's kind of the premise and the difference between these proprietary and open source models. But here's why I think it's interesting because Inkling, where it came at and the uh where it landed on the artificial intelligence index, uh 41. So right now your uh leaders are uh Fable 5 with a 60 and GPT-5.6 soul with a 59. All right, so a 41, you know, seems like okay, that's a pretty big drop-off, middle of the pack, right? True, but when you put into context that, you know, go back about seven-ish months, uh, the leaders of the pack at that time, well, it was GPT-5.2 uh with a 42. So uh the numbers change, all right, because the benchmarks that go into this, uh the artificial uh artificial analysis intelligence index, those uh benchmarks get updated. But it's essentially, you know, about a dozen or so different benchmarks that are always updated and rotated, uh, that tell you how good is a model compared to you know the most important factors. And Inkling, again, only being about seven months behind the frontier is actually pretty impressive for the first release from a company that we didn't really know what they were working on, right? We really didn't start hearing from Thinking Machines Lab for their first like year that they launched, right? So they launched, I think it was uh quarter one or quarter two of 2025. We didn't really hear anything from them for a year, and then we heard they're working on you know Tinker, and then we heard that they were making their own model for fine-tuning, and then we saw this um you know, this bi-directional voice model, but we didn't actually see the inkling model until well, less than 24 hours ago. But if you put it like that, it is about seven months behind the frontier, but it is an American model, which is actually important, and it's multimodal. Those two things alone, I think, have the potential to reshape what's possible. Because my thought is right, that model right there, it's not gonna, you know, if your team is AI native, if you have, you know, agents running, if you have workflow set, you're not gonna be able to swip. Like, let me just be honest, right? You can't just, you know, click copy and paste uh and put inkling in there. But for those companies that are still finding their footing or larger enterprise companies that are looking to uh, you know, chunk off a big piece of their workflows to something maybe more affordable. Inkling is actually not a bad option, uh right. So the real product though is Tinker. It is they are trying to turn fine-tuning into a service. So Inkling, yes, it can be downloaded, but even those compressed versions need 600 gigabytes of memory. So yeah, you're probably not running this or using this unless you are a large enterprise organization uh with your own, you know, GPU server infrastructure. So Tinker, though, manages the training so companies can customize the models without actually owning the GPUs, and the business model essentially turns that openness into what I think could be the first strategic reset. So uh let's talk about those potential strategic resets. Uh, reset number one, American open weights could reopen procurement. So many big enterprises haven't been able to touch some of these open source Chinese models. Well, be because of the current, you know, China-US relationship, right? Especially for those companies that do business with the government, right? And we've seen the US government get uh much more involved here recently, um between you know, export controls, um, but also a huge thing here is that we've also seen reports in the past week that China may actually, which is, I don't know, funny or interesting, right? That China may shut down uh its models to other countries, right? So even though they are distilling from US companies, we've seen reports they may not let, you know, who knows how that uh will be set up, but they may not quote unquote allow overseas companies to use their open source models. So I don't know how open source that actually makes them, and especially if they're just distilling them from US labs, anyways, but that's beside the points. But I think for so many enterprises, they haven't been able to look at the open source category yet because of that reason, right? If you're a company that has big government contracts, if you are using uh Chinese open source models, that's gonna put you in a sticky situation, or your uh RFP might be DOA, right? You might not even be considered if you are a company that has been using uh Chinese open source models. So that's a big unlock. All right. And I think that many companies have also just wanted to use open source models, but the whole fact that these are Chinese models, and you know, we've seen reports on, you know, can you actually trust, right? Like what's kind of the uh the the messaging that may be coming out of this that may not be in line with companies that have stronger, you know, American values or stronger Western values. I'm not gonna get into that, but you know, there's obviously many different reasons, not just geopolitical reasons, that so many companies here in the US haven't really been able to uh, you know, convince their board or convince, you know, anyone to go down the open source route just because it is all Chinese models. So here's potential reset too. Um, the model overhang makes this customization very timely. Uh so I've talked about this a lot over the past like three months, but I think we're now at this point where there's a model overhang. Um, it's a little bit different than the capabilities gap that I talked about. There's the great anthropic study uh from uh it seems like it was from so long ago, but it was only from a couple of months ago, their uh labor index report, uh, you know, that essentially showed the capabilities um of these models and then what companies were actually using them for, right? They uh anonymized, I think it was 400,000 uh agentic chats and they mapped it all out. And essentially what they said is these models are so capable, but you know, maybe you know, enterprises are using, you know, on average, about 10 to 20 percent of the model capabilities. So essentially these models right now, the frontier models, are way more powerful than most than the average company actually needs, or even has the actual capabilities to take advantage of. So that kind of adoption gap makes fine-tuning middle-of-the-pack models for stable repeated work, maybe an actual new and intriguing area of AI. You know, couple that with the fact that we have seen this whiplash, the um token maxing to you know, value maxing or token efficiency whiplash, where you know, earlier in, you know, from December 2025 to I would say March 2026, you know, enterprise companies were like, yes, you know, we've been all in on AI, so go use as many tokens as you can, right? And then it's like, wait, these token costs are getting higher and higher. And, you know, certain uh model providers aren't token efficient, right? Uh the data says that is anthropic, right? Um, so all of a sudden these companies have these huge API bills. So now we've seen uh, you know, in the May, June, July, this whiplash of companies being like, wait, we have to start reining spend in. And part of it is, well, they're realizing that we don't need, you know, a fable five type model for you know a hundred employees rewriting their emails. Uh, and that's what I think these middle-of-the-pack bespoke customized models could actually be a big thing. And why is that? Well, fine-tuning could return because now we actually have frontier models that are smart enough to teach, right? Which we haven't really had before. I actually have you know uh a couple exciting examples that I'm gonna go over here. But you know, this the concept of fine-tuning, right? So this is the teacher-student model. This is essentially right, and we've also got reports. If you read our newsletter every single day, we've gotten now some sniffs of real, like, okay, we're at this point of recursive self-improvement where you know models are actually, according to research, right? There's actually a very interesting paper from earlier this week that, yeah, the models are actually improving themselves. Um, right, but that also lends itself to the big, big models. Um, I think maybe GBD56 Soul is the first one I've seen uh in mass that we've seen actual real examples from uh that can teach and train smaller, much, much, much smaller versions of themselves to be used for many different purposes, right? Because these frontier models can generate data evaluations, code, foul uh failure analysis uh analysis. So I think it is not just you know, tinker this fine-tuning as a service from thinking machines and uh their new model. It's the combination of that plus this new tier of frontier models that make fine-tuning a reality, right? Because it has to become common language, right? And I think we're gonna see that. It's like I could right now, and I have an example that I think is a really awesome one, uh, but I think I forgot to include the the photo of it in my slide. Um, but you can, if you have a decent enough machine, uh, even a local one, right? Like a Mac Studio, you can start building very, very small models yourself with hardly any idea of what the heck you're doing for very small purposes. All right. So three of these more recent examples of this concept of, you know, well, in this case, GPT-5.Soul uh being in the fine-tuned models. So OpenAI's uh Jason Luai uh said that GPT-56 Soul actually helped post-train OpenAI's small model, GPT-5. Completing the work estimated to require two researchers, roughly two weeks. Uh, this is the one I forgot to put the screenshot in uh on my on my slides here. Uh, but Pietro uh Sherano, hopefully I got the name right. Uh so he's uh an CEO of uh of an AI company, but he said just a fun little project that he worked on. He used GPT-5.6 to build a small local model from his iMessage history. So essentially, I think he had it, you know, work overnight or something like that. And it literally read every single message in his uh Mac history. Uh, you know, so on your Mac you can access your iMessage. If you're you know one of our green bubble friends, you're like, what does that mean? Right. So on my computer, I have my text messages. So he just let it go. It read everything, it understood everything, ran some models, some evaluations, and all of a sudden he had a literal uh small language model that GBT56 made for him, trained on that. So this isn't like a custom GPT, right? It's like, oh, I'm using the big model uh to um give it instructions. No, it is a separate model itself, right? Absolutely crazy. Uh then NVIDIA used Codacs. They just put a blog post out uh about this, I think on Tuesday. Uh NVIDIA just used Kodaks and GVD5.6 to post train its Cosmos 3 nano model, reportedly improving accuracy in it from 454 to 93 in one day using two prompts. All right, I'll I'll I'll make sure to. Reshare that in today's uh newsletter. I think we did put it in Tuesdays or yesterday's, but uh since I'm mentioning it on the show, I'll make sure to put it in there. But here's the reality two prompts, these are off the shelf creating or improving existing models with today's frontier technology. So it's not just inkling, right? It is more of a reflection of the combination of these very powerful frontier models that can literally build, run the evaluations, run the testing, run the QA, can literally build large small language models on their own, uh, and also uh train you know other variations of themselves. So, one example though, that we got from Thinking Machines, they did share uh recently one use case about Bridgewater, right? So using their Tinker service, right? So this fine-tuning as a service, uh Bridgewater customized Quen, uh, which is a Chinese open source model. So this is, you know, obviously before Inkling was uh available, or at least before they were ready to use it and talk about it, uh, they did come out with this customer use case of Bridgewater customizing uh the Chinese open source Quen model through uh Thinking Machines Tinker platform for recurring financial judgment tasks. So it report uh it reportedly beat the best tested frontier model while costing 13.8 times less. So in this case, specialized judgment beat the frontier defaults, and that kind of explains this potential model shopping shift. So let me just put it out here, y'all. Not one to be like, hey, I told you guys this a long time ago. Maybe I was just a little too early or a little too weird, but I think two years ago in my AI prediction series, right? I talked about this very thing and the uh the rise of potentially seeing thousands of small language models created by the large language models. And although we're still maybe not there because uh, you know, there is this whole thing called, you know, compute is scarce, uh, right? Where you necessarily you can't have uh anthropic and open AI and Google and Microsoft and Meta and Grok. They can't necessarily, you know, pull uh gigawatts of compute to create thousands of these bespoke uh middle tier and small language models, but that's still the reality. And I think that this here, uh, the combination of a GPT-5.6 soul uh being able to create small language models and uh post-train its own versions and the combination of thinking machines uh coming up with this as a service. I think there's finally my uh I forgot if it was late 2023 or late 2024 prediction could be coming true. Where I think we are eventually and very soon going to see large language models create hundreds or thousands of versions of small language models uh because they're cheaper. And there's finally the appetite to care about it because nine months ago, people didn't necessarily care because we were still right, we were still on this, you know, this richie rich blank check AI, right? It for $20 a month. You couldn't, most 90% of employees, right? If you had a team or a business plan paying $20 to $50 a month per seat, most employees couldn't get through that. So companies didn't necessarily care about being efficient with their AI. The fact of right, the the concept of a small language model or a a medium-sized model fine-tuned for specific tasks didn't necessarily matter because the big model could still do it. And everyone essentially, we were playing with monopoly money until about four months ago. So the whole concept of small language models or fine-tuned models literally didn't matter because it was free money. Now, because of the whiplash, this is more important than ever. So, as we wrap, here's what I want to talk about. Reset free, right? Model shopping is now big. And this is now, I think, part of the executive playbook. And we saw Microsoft, right? Microsoft is reportedly gonna start moving uh away from open AI and anthropic models, not moving away, but they're gonna start mixing in their own models as well as reportedly deep seek models. So even the biggest enterprise customers are starting to realize that for a majority of day-to-day tasks, you might not need the number one model in the world to do a big chunk of the work. And I think now it's more of this concept of maybe renting frontier intelligence for those uh, you know, for the ambiguous and risky uh and and changing work. But I think the future, which I've talked about, is using the right model for the right purpose at the right time. And eventually that will become automatic, right? I've actually built some things like that. I'm like, I'm like, I wonder why no one else is doing this because it's not like especially since GPT 5.6 came out. Like I built skills that essentially model route, right? I can use Fable inside of Codex, I can use GPT 5.6 uh inside of Claude Desktop, right? If you have a little bit of skill and enough patience, um, you can do those things, right? And I've done the same things where um, not that I ever hit my usage, but just a practice, right? Where I have I've built in model routing within codecs or you know, chat GPT work. So if you are a cost conscious um, you know, company, this is the future of working with multiple models, and you're not just gonna throw, you know, every single request at one big model. So here's the new AI playbook. You own the workflow, but you're probably just gonna rent the frontier or just exclusively use the frontier for those specific purposes. So you have to score every workflow by the stakes, the volume, the privacy, and how much proprietary judgment actually differentiates it. So you have to start to default to more of this economical model, right? Like you have to start mapping out that workflow and saying, okay, maybe we send the 20% to you know, if you especially if you're using it via the API, to that frontier model, right? But I mean, my gosh, talk about middle, middle and lower tier models. I mean, open AIs, uh, Terra and Luna, when it comes to a cost efficiency, are legit off the charts, right? So if I'm advising companies, you know, I'm saying like you should be using these models for the most part, right? The the the Luna model on like max setting is probably more than enough for almost 90% of the work that you would do. So it is kind of shifting away from this one model for all purposes to understanding, which I know is is confusing because it is easy to hit that easy button, but we can't just hit that easy button every single time because that button is going to start to cost more and more money as it requires more and more power. Uh, right. So you have to default to those sometimes more economical models, route by difficulty, and then fine-tune potentially stable, repeated, measurable work. And you have to stop asking which model is the smartest and start asking which model is enough for each job. All right, that's a wrap. I hope this one was helpful. The pretty exciting release, maybe not just for the model itself, but more for what it represents. So I hope this was helpful. If so, please let me know about it. Go sign up for our free daily newsletter. Drop me a line when you get that automated uh, you know, email, uh, welcome email. But also, if you could do me a favor, go subscribe to the podcast on Spotify or Apple Podcasts. Appreciate you tuning in. We'll see you back tomorrow and every day for more everyday AI. Thanks, y'all.