ChatGPT Work launches with agentic AI capabilities as writers share practical tips for using ChatGPT

How to use ChatGPT Work - and my top 10 tips for getting started with agentic AI

Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways * ChatGPT Work can browse, manage files, and complete tasks. * Free users can try it, but paid plans offer more capacity. * Start small, limit permissions, and double-check its work. ChatGPT answers questions. ChatGPT Work does work. That's because ChatGPT Work is agentic, which is a fancy way of describing how some AIs can go off and accomplish a series of tasks. Because ChatGPT has become more capable, the differences between ChatGPT and ChatGPT Work can be subtle. For example, if you ask ChatGPT to plan a travel itinerary, it will search the web and gather together a list of great destinations. That's the AI doing something. Also: 'I can't stop': 80% of developers find AI coding more addictive than helpful It gets even more blurry with ChatGPT's Deep Research mode. Deep Research is essentially a special-purpose agentic web researcher. Deep Research can dig around the web and build a much deeper analysis than ChatGPT. ChatGPT Work can do all that, and more. It can open your own Web browser, or its own internal one. It can move around the web on screen (rather than or in addition to retrieving a web page's HTML data). It can create and move files. It can process through a long sequence of steps. This approach is very much like what an AI agentic programming tool like Claude Code or OpenAI's Codex can do. In fact, agentic work tools like Claude Cowork and ChatGPT Work evolved directly out of agentic coding tools that some users decided to use for non-coding tasks. Also: This AI expert says the job apocalypse isn't coming, even if you're a coder - here's why Following a surprising usage surge in non-programmer use of those programmer tools, both Anthropic with Claude and OpenAI with ChatGPT launched agentic work products. Today, we'll investigate starting with ChatGPT Work. What you need for ChatGPT Work ChatGPT Work is available for all ChatGPT users, even those on the free tier. There are limits, though. Agentic operations take a lot of AI juice (typically measured in tokens). Free users are given access to only a small amount of capacity. As you pay more, you get more usage capacity. I have been using the $20/mo ChatGPT Plus, which has proven to be pretty capable. When I use Codex for coding, it tends to slam shut after about five hours of intense work. Still, that's five hours of intense work for $20. That usage limit resets after a while. I've found I can get about 20 or so hours of agentic coding done each month on that tier. ChatGPT Work uses a similar amount of capacity. You can probably do a fairly large project with Work every few days and not feel all that limited on the $20/mo plan. Unfortunately, OpenAI (and this is true of all the AI companies) doesn't have a single metric that you can measure by. Each request requires a different mix of resources, so an absolute usage capacity isn't really possible. But I'd think of it this way: free gets you a taste, $20/mo gets you fairly regular moderate usage, and the higher tiers let you binge. Also: 75% of developers I surveyed prefer Claude Code - here's why they choose it over Codex Beyond your account, you'll need to either sign in to the ChatGPT web page or download the desktop app. For any tasks you want to run on your computer, you'll need the desktop app. Note that up until mid-July, the ChatGPT desktop app was a chat app. Then, in July, OpenAI combined ChatGPT, Work, and Codex into a "super app," which had the unfortunate side-effect of killing some of the ChatGPT app's best features. So now, if you want to use ChatGPT Work on the desktop, download the ChatGPT desktop app. Also: I loved ChatGPT Desktop until OpenAI gutted it to make room for Codex and Work You can ignore my grumbling about the removed features. OpenAI still hasn't bothered to put them back in. But I'm not bitter. It's been almost two months. I'm not holding a grudge. Not at all. How to use ChatGPT Work Once you launch ChatGPT using either the desktop app or in your browser, you'll see two buttons above the chat message bar. One is Chat, and one is Work. Clicking the buttons (above) changes the mode. On the top, we're in Chat mode. On the bottom, we're in Work mode. Notice also the option to choose the folder you're working in on the bottom. This approach means you can restrict a complex work project to a single folder or hierarchy. Also: A low-tech solution from the past may be your best defense against AI deepfakes Now, all you do is prompt the AI just like you always have. The difference is that Work can do a lot more. Last month, for example, I had ChatGPT Work dig through a large collection of PDF files, rename them, and organize them in topics. You can read about how I approached this project here. One thing that did make me nervous is that Work hasn't been asking for permissions as often as I'd like. I've generally learned to tune my requests so it doesn't have a reason to run amok, but I would prefer to be asked for an okay more often. An example research project Before I set you loose on ChatGPT Work, I'd like to show you a research project I gave the tool. I'm planning an article comparing vertical mice, and I want to test a bunch of them to see which works best. But I don't just want vertical mice. I want wireless vertical mice that can switch between two to three devices (three preferred). I started digging through Amazon and realized I'd spend an hour or more trawling specs because the product headline doesn't always indicate how many devices a mouse supports. Also: You can earn your Google Al Professional Certificate for free - and I highly recommend it Then I remembered that I have an agentic AI at my disposal. It was time to delegate. I gave ChatGPT Work one prompt that consisted of the following components: I want you to do some Amazon research for me. I want you to find wireless vertical mice. That was my basic overview. You're going to research mice on Amazon. Then I discussed the key restriction that would narrow the search and save me time: They must be able to switch between at least two Bluetooth wireless devices (but I prefer three). Ignore vertical mice that aren't wireless and ignore vertical mice that only support one computer. Finally, I gave it an assignment only ChatGPT Work (and not ChatGPT or ChatGPT Deep Research) could do. I told it to check each candidate for Prime shipping and fairly immediate availability, and then open each one in its own tab. I wanted to quickly toggle between tabs, close ones I was uninterested in, and then narrow my evaluation scope: When you find them, open each candidate in its own tab. Only choose mice that are Prime and will ship in less than a week. As you can see in the image below, that process took ChatGPT a little over 10 minutes, during which I went and got myself a nice cup of coffee and played with the dog: Ten quick tips Here are some tips I recommend, based on my experience: Well, there you go. I've found that I'm using ChatGPT Work and Claude Cowork more and more. I switch between the two mostly so I can share my experiences with you. Frankly, my experience has been that both are pretty much on par. And with that, let's call it a day. Have you used an agentic AI coworker like ChatGPT Work? Let us know in the comments below. 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, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.

The best way to use ChatGPT for writing isn't writing -- 7 prompts I use to make it my creative partner

AI shouldn't replace your voice. Here are 7 copy-paste prompts to pressure-test ideas, fix pacing, and catch bad habits before you publish I write for a living and I use AI every single day. But that doesn't mean that I use AI to write. I've loved the art of writing long before artificial intelligence was available for use by the general public, so the idea of using it to do what I've enjoyed doing for decades really doesn't make sense to me. But I have found that some of the best ways to use AI do support making my writing in general -- just not generating the actual words. Instead, I use it before I write, while I'm researching and after I've finished a draft. Sometimes I use it when I know something isn't working but can't quite figure out why. Using AI has become my writing partner as if I was in a writers' room. Since I work remotely, leaning on it has become a way to ask unlimited questions, including annoying ones, that might otherwise annoy an actual human writer. If you write for work, school or just yourself, and have questioned whether to use an AI humanizer to make your AI writing less detectable, scrap your process. Here are seven ways I'd use ChatGPT to make the process easier without asking it to write a single sentence for you. 1. Pressure-test your idea before you start One of my favorite things to do with ChatGPT happens before I've written anything. I'll explain what I want to write about and then ask it to poke holes in the idea. I'll give AI my "big idea" and ask what it thinks. Now that ChatGPT is less people-pleasing, it actually tells me when my ideas aren't as great as I think they are. And when the ideas don't have legs, AI offers suggestions to help make them work. Prompt to try: I'm thinking about writing an article for my Substack arguing that [idea]. Don't help me write it. Tell me what's weak about the premise, what I'm assuming and what a skeptical reader might push back on. The important part here is telling ChatGPT what you don't want. Ask "What do you think of this idea?" and there's a decent chance you'll get an enthusiastic response about why it's interesting. Asking it specifically to find problems tends to produce something much more useful. You might ignore half of what it says. That's fine. The point isn't to let ChatGPT decide whether your idea is good. It's to uncover the questions you need to answer before you start writing. 2. Find the questions you haven't thought to ask This is one of the most useful ways I've found to use AI for research. Once I know the basics of a subject, I'll tell ChatGPT what I've learned and ask what I'm missing. Prompt to try: Here's what I know about [topic]. What questions would you want answered before you felt like you fully understood this story? Don't answer the questions. Just tell me what I should investigate. That last instruction makes the biggest difference for me every time. AI models can confidently give you an answer when what you really need is a research direction. I don't necessarily want ChatGPT telling me the answer. I want it pointing at the hole. This can be particularly useful before an interview or before sending out a pitch to a publisher who might publish your book. You can give ChatGPT the questions you've already prepared and ask what you haven't asked. I've had moments where one of those suggestions made me realize there was a much more interesting story sitting underneath the one I thought I was reporting. 3. Use AI to understand your research, not replace it There's a huge difference between using ChatGPT as a source and using it to help you understand your sources. I wouldn't cite ChatGPT for an important factual claim. I will absolutely use it when I'm staring at technical documentation and thinking, "What on earth does this mean?" Or, if I hear a "fact" that I'm questioning, I will ask ChatGPT to find the source. Prompt to try: [I'll paste in the relevant material and ask questions about it] Explain this to me like I'm intelligent but completely new to this subject. Stick only to the information I've provided. Tell me if my question can't be answered from it. Then I go back to the original material. This is especially useful when you're entering an unfamiliar subject and don't yet know the vocabulary. Once you understand a little about something, be it a historical event or a gadget, you suddenly know what to search for, who to talk to and what questions to ask. The important part is that ChatGPT hasn't done the research for you. It has helped you get better at doing the research yourself. 4. Ask ChatGPT to disagree with you This might be my favorite one. AI assistants have a tendency to be agreeable. That's nice when you're chatting with them. It's considerably less useful when you're trying to figure out whether an argument holds up. So don't ask: Does my argument make sense? Prompt to try: Make the strongest possible argument against my position. What evidence or reasoning would someone use to challenge it? You can go further: Assume you're a skeptical editor/publisher/manager/professor who doesn't think this story/presentation/essay should be published. Make the case against it. Now you have something to respond to. Maybe ChatGPT's objections are terrible. Maybe you've already addressed them. Or maybe it identifies the exact thing that's been bothering you about the piece but you couldn't put your finger on. Either way, you've forced yourself to defend the argument rather than simply asking an AI to validate it. 5. Ask where your draft gets boring To be fair, I usually know when something is getting boring. But there are a lot of questions you can ask AI to avoid boring your audience. Once I've written something, I don't necessarily want ChatGPT to rewrite it. Sometimes I just want to know where it drags. Prompt to try: Read this as an impatient reader. Tell me the three places where you're most likely to stop reading. Don't rewrite anything. Explain why you lost interest. I love this because it separates diagnosis from treatment. ChatGPT can tell you that you've spent four paragraphs explaining something the reader probably understood after two. It can point out that the opening moves slowly or that an interesting idea appears halfway through the article when it probably deserves more attention. But then you fix it. That means the voice, rhythm and decisions stay yours. AI simply acted as an editor. 6. Let ChatGPT find your bad habits Every writer has bad habits. I know mine is using "because" a little too often. I also often use the same transitions. But whether it's that every third paragraph starts with "But" or you have a favorite word that appears 14 times in 800 words, it's not always obvious. After you've been starting at the same draft for an hour or days, those things become invisible. This is where AI is fantastic. Prompt to try: Don't rewrite this. Analyze my writing habits. Look for repeated words, phrases, sentence structures, unnecessary qualifiers, clichés and patterns I'm relying on too often. You can even make this more useful over time by giving ChatGPT several pieces you've written and asking what patterns appear across all of them. Some of the feedback will be subjective. You don't have to accept it. But once someone, even an AI someone, points out that you've used the same phrase six times, it's very difficult to unsee it. 7. Give your finished draft one last interrogation This is the step I'd do immediately before considering something finished. Don't ask ChatGPT to polish the draft. Don't ask it to "make this better." Ask it to interrogate it. Prompt to try: Act as a skeptical final editor. Don't rewrite anything. Identify factual claims I should verify, assumptions I'm making, contradictions or inconsistencies, places where I haven't supported my argument and questions a reader may still have when they finish. For something important, I'd make the instruction even stricter: Don't tell me what's good about the piece. I'm only interested in potential problems. Then work through the response yourself. ChatGPT might flag something that doesn't need fixing or it could misunderstand your argument, it's just a way to have another "set of eyes" take a pass at your work. Bottom line: AI doesn't have to be the writer There's an understandable fear around using AI for writing: If I hand too much of the process to ChatGPT, at what point does the writing stop being mine? My solution is simply to not hand over the actual writing to AI at all. Instead, use it as a supportive writing partner. Use AI to question your premise, poke holes in your research, disagree with you and find the paragraph where you started rambling and point out that you've used "actually" nine times. Then close the chatbot and fix those things yourself. Because the best use I've found for ChatGPT isn't asking it what I should say, but how to question what I've said to make the final draft worth publishing. Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok.