Vstock LLC/Getty Images
It's only been almost three years since generative artificial intelligence (AI) hit the mainstream as a new paradigm of productivity, but here we are -- it's everywhere.
I test AI tools as part of my work. I'll dig into just about any AI-related technology and see what I can make it do. Many of you have read my ongoing shootouts comparing AIs for programming and AI content checkers, among other kinds of tools.
But that's using AI in a rigorous lab environment to provide test results to ZDNET readers. Like many of you, I've also started using AI to augment my workflow and increase my productivity.
Also: The best AI for coding in 2025 (and what not to use)
I wear a lot of hats; I run a small business with my wife, who also has her own business, where I'm the tech guy and designer. I also work with a number of industry groups. I have a fairly popular security software product for WordPress users. And I'm constantly working on projects, ranging from 3D printing the ultimate charging tower, to trying to make an AI-assisted Etsy store, to composing and publishing music, and using an AI for help with some of the marketing activities.
I should note that I never, ever use AI to produce my core content. No article, song, or social media post is ever written using an AI tool. My work product is mine. But I do use AI to help me get through other aspects of my workload.
I have a particular interest in how AI helps programming, how AI can support graphics work, and how AI can support video production.
Here are the tools I'm willing to pay for -- and why.
1. ChatGPT Plus - $20/mo
Speaking of AI and programming, it has essentially doubled my programming output. I use AI to help me with common-knowledge programming. I talked about it in-depth in my 25 tips article, but the core benefit is getting ChatGPT to write code for published APIs, so I don't have to spend time searching for code examples and trying to reverse-engineer comments on various programming boards.
And yes, I mentioned ChatGPT. While more chatbots capable of passing my programming tests have been introduced in the last year, ChatGPT does the job well enough, and hey, who wants another monthly fee?
Also: How ChatGPT actually works (and why it's been so game-changing)
In fact, that's a big part of why I'm paying $20/mo for ChatGPT Plus. Sure, I've signed up and paid for some of the other AIs just to test them, but ChatGPT Plus is the only chatbot I have found so consistently useful that I keep it as a regularly used tool.
I use ChatGPT for lots of research tasks, sometimes throwing math problems at it, and all sorts of other questions and problems I'm dealing with. While I never take its output as an unimpeachable source of truth, I do find ChatGPT to be a very useful sounding board, substantially more so than a quick Google search.
Now, to be fair, I did outline five ways that an AI could help me in Gmail. If Gemini could do these things reliably, I'd sign back up in a heartbeat. But I just don't need the current email message I'm reading summarized, and I sure don't need it to write a friendlier or more professional version of whatever I've currently written. I tried Gmail's new AI unsubscribe feature, and it only found about 10 newsletters, yet I get thousands of emails and hundreds of newsletter-style messages every day. So, I'm leaving that one unbought.
2. Midjourney - $10/mo
I played around a lot with DALL-E, ChatGPT's earlier image generation tool. But recently, OpenAI introduced a new image generator in GPT-4o, and it's quite the beast. I have found that it generates great results, but it has more guardrails than another tool I pay for, Midjourney.
Also: I tested 10 AI content detectors - and these 5 correctly identified AI text every time
But even though I get image generation with my $20/mo ChatGPT Plus fee, I pay an extra $10/mo for Midjourney. Why?
One of the reasons is subjective. I like a lot of the images I get with Midjourney. Midjourney also allows me to describe artist styles, and lets me riff off a vast array of stylistic choices. ChatGPT, perhaps because of guardrails imposed by OpenAI, doesn't present as many choices.
But I also have two specific and objective reasons for paying for Midjourney. First, because image generation is so subjective, it's nice to have a variety of tools when seeking a representation of what you have in your head. I'll try different prompts and even the same prompts with both tools and take what works best.
Also: How to selectively modify a Midjourney image to create an artistic statement
Second, every month I generate a promotional image for my wife's online business. She has an e-commerce site that supports a popular hobby. Each month, on her very active Facebook group, she gives a craft-along theme to her users. I generate an image for that theme. Over the months, I've found that Midjourney does a far better job of generating an image that incorporates elements of the hobby. That said, some months I bounce back and forth between both tools until I can get an image that meets her business's needs.
Because Midjourney shaves what used to be two to three hours of work pushing pixels in Photoshop to generate those images down to about 10 minutes, it's worth the $10/month to me just for that project.
Photoshop Generative Fill - Honorable mention
In the title of this article, I said I pay for two AI tools. That's sort of true. I pay for Adobe's Creative Cloud suite in addition to ChatGPT Plus and Midjourney. But since I've been using and paying for Creative Cloud -- and before that, Photoshop -- long before there was generative fill, I'm not counting it in my AI tools list.
Also: I use Photoshop's AI tool every day - here are my 5 essential tips for the best results
If Adobe removed generative fill tomorrow, I'd still pay for Photoshop. To be clear, I don't like paying for it. It's costly, and the two-computer license limitation is restrictive. But a few years back, I tried switching to Affinity Photo, which at the time was $50 (it's now $70). That one-time fee is roughly what I pay each month for Creative Cloud, so it had a lot of potential.
To be clear, Affinity Photo is a fine application. But I've been using Photoshop since before the Clinton administration. To say I have Photoshop muscle memory is an understatement. It's a product I use almost every day. Switching to another application, while I could do it if I had to, slows down my workflow considerably.
Also: What to do if Generative Fill is grayed out in Adobe Photoshop AI
So, I don't consider my monthly expense for Creative Cloud to be an AI expense. That said, I find generative fill (and its various other AI tricks) very helpful. I often use it in concert with Midjourney and ChatGPT image generation.
Three tools I'm thinking about
I run a business online and, as such, rely on a wide variety of cloud services. Those fees add up, and now they're all going up in price. So while it might be nice to add more AI tools, I'm keeping it under control. It's very easy to just click OK and find yourself spending hundreds of dollars more every month.
That said, I am thinking about adding three more tools. I'm a bit hesitant, because each one has its annoyances and limitations, but they're on the short list for a quick order if I can ever justify an immediate performance improvement on one project or another.
Notion AI
The first is Notion AI. I am deeply invested in Notion for all my project work. I also use it to write and organize all my articles, as well as schedule them, plan them, research them, and capture notes and assets. Notion AI is interesting because it would work like NotebookLM, limiting its knowledge base to my Notion account. That could be very useful as I work on more projects. But at one point, when Notion overcharged my wife's account, they were completely unsupportive and unsympathetic. So, I hesitate to give them more business.
NotebookLM Pro
Google's NotebookLM Pro is another contender. Now that Pocket, the article archiving service, is being discontinued, I considered using NotebookLM Pro as a replacement. The idea that I could save articles in NotebookLM as sources and then have the AI review them, summarize them, and analyze them seemed ideal, especially as a research tool.
But... the free version of NotebookLM only allows 50 sources per notebook. The Pro version, which is normally another $20/mo ( you can usually get a few starter months at a discounted rate), increases that limit, but only to 300 sources per notebook. My archive has well over 30,000 sources, which is beyond NotebookLM's limits. There is a $249/month plan (yowzah!), but all Google will say about limits is "Highest limits and best model capabilities (later this year)". What does that even mean?
Descript
Descript (for $16-$24/mo) is an AI video editing tool. This isn't a tool that does text-to-video generation. Instead, it's a tool that helps you take your video clips and edit them. Right now, I'm a very big Final Cut Pro user. Final Cut has added some AI features, but it lags far behind DaVinci Pro and Premiere Pro (because Apple lagging in AI is no surprise, right?).
Also: How to use ChatGPT to write code - and my top trick for debugging what it generates
Descript automatically removes filler words and retakes, cleans up sound quality without any fuss, and does automatic multicam editing. It also promises to take long-form videos and automatically create clip videos, which could be a huge time-saver. The product also has some more "out there" features which I wouldn't use, including fake avatar generation and fake speech generation.
The thing is, Descript is aimed more at multiple talking head videos. I'm not sure it could handle the sort of in-depth technical hands-on project videos I do. So, it's still in the "maybe someday" category, for now at least.
What do you use?
Do you pay for any AI tools? Which ones, and why? Is there an AI tool that you strongly recommend I should be using that I didn't mention? Feel free to answer these questions and let us know your thoughts on AI subscriptions in the comments below.
Want more stories about AI? Sign up for Innovation, our weekly newsletter.
You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, and on YouTube at YouTube.com/DavidGewirtzTV.