Nate B Jones walks through his complete personal AI tool stack for November 2025
A solo presenter shares which AI tools he uses, what he uses each one for, and where each one falls short.
Summary
Nate B Jones of AI News & Strategy Daily walks through his complete personal AI tool stack as of November 2025, covering tools for analysis, writing, PowerPoint, Excel, search, web browsing, and command-line coding. He is candid about the limitations of each tool — including context window problems with Claude, benchmarking he doesn't trust, and data sovereignty concerns with a Chinese open-source model. His central argument is that no single AI tool does everything well, and that building a stack means matching the right tool to the right task rather than defaulting to one platform for everything. He also emphasizes that AI-assisted writing still requires the author to own every word produced.
Key Takeaways
FULL TRANSCRIPT
Overview of the stack
Nate B Jones: Today I'm going to do something I've never done. I'm going to share my personal AI tool stack — all the way through, end to end: what I use things for, where they work, where they don't work, where I'm frustrated, and where I'm working around things, so that you get a sense of how this works.
ChatGPT — analysis and thinking
Number one: ChatGPT. Everyone uses ChatGPT. What do I use it for? I use it for analysis specifically. If I need something that has some memory, that can handle a lot of context, where I don't run out of context window, and I need it to think clearly — not for writing, for thinking, for getting the idea right, for being a thought partner back and forth — ChatGPT's thinking mode is very, very useful.
What's interesting is that the auto mode, the fast mode, is becoming increasingly useful for digesting large amounts of context quickly. So if I need a rough pass, I can do that. But I do not use it for writing. I find that ChatGPT can be used for writing if I push it, but almost always I go somewhere else, because the default voice for ChatGPT is not good enough.
I also don't use ChatGPT right now for PowerPoint, and I don't use it for Excel. I don't find that the finished quality of work is really there. I will use it to produce CSV files if it's just a simple one-sheet spreadsheet and it's got to be a table — ChatGPT does great at that.
Claude Sonnet 4.5 — writing, Excel, and PowerPoint
Moving on. I use Claude Sonnet 4.5 a lot. I use it specifically for writing, because it's very, very good at picking up voice and at following my instructions and actually getting voice right — especially if I give it a sample with that voice. One of the things I like to do is ask it to brainstorm in my voice back and forth with me. How can I write this better? Can I tweak this paragraph? Can I do a better job formulating this point? It's a writing assistant. It doesn't just produce it for me and I walk away.
I keep emphasizing this with writing: you are accountable for every word you write. So if you're going to put something out there, you better own it — however you made it, whether you're writing with AI or without AI. I find it's a great thought partner when I work with it that way.
Claude Sonnet 4.5 is also what I go to when I'm doing Excel analysis right now. It's got a great toolset for that. I have a whole guide out on that, and I find it super useful because I can actually dig in and understand what's in the data and produce a useful analysis or edit an existing file. All of those things are possible with Claude.
Kimi K2 — PowerPoint with caveats
That being said, there's an even better tool out there for PowerPoint that just came out recently, and I'm really torn about it. I'm going to tell you about it and give you the caveats, because you're getting a tour of the workshop — you get to see how Nate does his personal stack.
Kimi K2 is an open-source model out of China, and they have done a phenomenal job of putting together a PowerPoint skill that enables you to make useful PowerPoint presentations that look really good. I'll show one with the Substack write-up — you can see it's a very simple prompt, but you get a very useful PowerPoint out of it.
That being said, because of where the data is located, you cannot really use this in the US or the EU for corporate data. The protections just aren't there. So if you're doing a fairly generic presentation where you're using publicly available information off the internet, it's phenomenal. If you're using it for personal use and you don't mind, it's phenomenal. But if you need corporate data protections, I find that Claude Sonnet 4.5 is super, super useful for PowerPoint creation — it's just a little bit less designed.
I tend to prefer a spare, elegant, minimalist approach with my PowerPoints. It's going to be what you need it to be with Claude — it's going to be usable, it's going to be presentable. But if you have a very elaborate PowerPoint style, I've got to be honest with you: you're not going to get what you want out of Claude. You're not going to get what you want out of any AI right now. The AIs that are doing the work with PowerPoint either have fairly strong preset styling or they're very clean and minimalist and elegant.
One thing I will call out: I do not use Kimi K2 for thinking. Despite the benchmarks it recently took, I find that it's just not as useful in practice. This is why you don't trust benchmarking. Instead, I use ChatGPT for thinking, and I can pull outlines and put them into Kimi K2 if the data is not sensitive.
Word documents and Claude's context window limitation
Sonnet 4.5 is also surprisingly useful for formatting Word documents. You just have to ask it and be specific about what you want. That's another useful tip.
The weakness — and I know people are going to call this out, so I'll just say it — I struggle with it too. Claude has context window limitations. In particular, if you are building a lengthy Excel or PowerPoint skill, you will struggle at some point with Claude running out of context.
How do I deal with that? I wrote in my guide on PowerPoints that you want to think about chunking the deck. You don't necessarily want to generate the whole deck in Claude at once. You want to break it up into pieces — maybe five or six or eight slides each. You also want to separate out the data and the narrative piece on bigger decks so that you can get those right beforehand and you don't burn tokens on those in a deck-creation conversation.
If I run into issues once in the chat, I always go back and restart the chat with a smaller ask. I have very little patience for running into issues more than once. So if I run into context window issues on Claude, I'm always going to condense the ask down and go piece by piece. I will find out where the context window runs out and I will get usable information in the meantime. Do I start again? I absolutely do, but I'm very careful not to repeat my mistake. I do not want to go back and hit the context window multiple times. So if I ask for a PowerPoint and it hits a wall at the end of the context window, I never repeat that ask. Same with Excel.
Search — Perplexity and Grok
What about search? I love Perplexity. I find that Perplexity is really useful for most general-purpose searches. I love the research piece, and I love the Labs piece where you can dive in on discovery and understanding and creating reports — it's super fun. But it's not perfect. There is one use case where it really struggles, and that is finding recent information on social networks about a trending topic. It's just not as good there.
And so, as funny as it sounds, Grok is really, really good at just that piece. I will go and say, "Tell me what people are saying on Reddit, on X, about this particular issue" — especially if you have an AI topic that is trending, or a brand-new product launch, and you want to know what people are saying about it. Grok is super, super useful for digging in and understanding that social conversation. I don't necessarily trust it for larger-scale thinking. I don't trust it for outlining. I don't trust it for general web research. But for finding social conversation on a specific trending topic, it is very, very good.
Web browsers — Comet and Atlas
What about web browsers? I'm really torn on this one. My general-purpose web browser remains Comet. I use it a lot. I like the data ins and outs it has. I'm particularly fond of its ability to do generative UI — so if I'm sending a message in LinkedIn, it will literally generate a message pane for me and let me approve a message before I send it. That's great. It has data ins and outs to a few places, including LinkedIn, where it's very easy for me to not have to go to the site directly. I've got to be honest: I don't love LinkedIn. I don't like to be there. Having a data in and out where I don't have to interact with the site is fantastic for me.
Beyond that, it's very useful because it combines Perplexity's search powers with an agentic browser. So it's easy for me to get a chat next to the browser, to understand what a page is doing, to ask it to do other things off-page for me — which I find really useful — while keeping the power of Perplexity in the meantime. I like the idea of the agent going off and doing things for me, and Comet really fulfills that.
I think the Atlas use case is really interesting. Atlas came out recently. I have found things I can do in Atlas that don't work as well anywhere else, but I feel like I am still figuring out where those use cases are. One of the big differences is that this is a ChatGPT-first browser. It brings my memories in. It is ChatGPT-first from a search perspective. It is ChatGPT-first from an engineering perspective. So you get all the strengths that come with that — the model remembers you, the model talks to you like it knows you, the model understands your preferences. And the more you use the model, the more it understands how you think about the internet.
This is all good, but it also means that you're going through ChatGPT for search versus Google — and that's a design choice they've decided to make. It makes sense from a product perspective, but it's something you have to decide if you want to live with. It also means that you are somewhat committed to their vision of the agentic future, which is frankly a little bit more buttoned-down and safety-first. They are keeping the AI agent on the tab with you for now. You're keeping an eye on it for sensitive tasks. They have walled off certain tasks — things around banking, etc. — that make sense to me. I wouldn't want it to be able to do that initially. So they're assuming that the AI agent is not entirely trustworthy yet and designing around that, which I appreciate.
One last thing to call out on Atlas: it has the ChatGPT brain for code, and that makes it super useful if you are trying to understand how to build something or how to review something. I used Atlas to look at a GitHub repo, and that was super useful — I could pull out a lot of the details of the code and get it into a format I could understand and process really quickly. I also used it when I was looking at Lovable. If you want to drive a build in Lovable using Atlas autonomously, you can do that, which is really cool.
Command line — Claude Code and Codex
Now we're going to come to the terminal and the command line. You thought I was done — I'm not done yet.
I love both Claude Code and Codex. I find that Codex is an extraordinary strategic thinker in the command line. I go to Codex almost daily when I'm trying to think through a problem and I need succinct, clear, strategic analysis. Codex is also really good at finding and fixing bugs. I love that I can throw a really messy repo at it and it just digs in and finds them.
I love Claude Code because it is so rich as an ecosystem. I can tie in cloud skills. I can tie in my local files. I love that I can tie in the MCP servers that I want to. And I love that — for lack of a better term — it has a friendly feel. It goes and does tasks and checks back in.
That being said, Claude Code has a very strong bias for action, and Codex is more thoughtful before engaging. You have to know which one you're going to choose, because otherwise Claude Code is going to be very tempted to just run.
So that's my personal stack.