Think of Dots as Codex, but for regular people. OpenAI’s new agent platform, called Dots, is full of cute little guys who can do your bidding. However, unlike the ultra-approachable Meta Muse, Dots feel very much like using workplace software that happens to be able to order you a burrito, with an emphasis on work.
OpenAI announced Dots earlier this week. Like Muse, Dots have blobby, anthropomorphic avatars and customizable names. In the future, OpenAI says you will be able to have multiple Dots, but right now you get one. The author named theirs Dotty McDotface. The interface looks similar to Muse’s; you chat with the agent in one window and follow its work in another as it clicks around on a virtual machine. But Dots are less like personal shoppers and more like coworkers: business software that can use other software. Your Dot’s virtual machine can access apps like Blender and GIMP right off the bat. You can also give it access to your own computer through the desktop ChatGPT app, and you can call your Dot if you need to work through things out loud. OpenAI is offering Dots first to users on its highest-tier accounts, including the $100-per-month Pro account used to test it. While Muse and its buzzy competitor Instinct cost nothing—at least for now—they are expected to eventually serve ads. OpenAI’s gated rollout suggests the product is targeted at enterprise users, despite the cute avatars.
The author tested other AI agents with day-to-day personal tasks, resulting in mixed results, and was curious how OpenAI’s bot would compare. Despite its enterprise vibes, OpenAI says Dots “can do nearly anything” with its cloud computer and access to apps. To start, the author had Dot try to schedule an installation appointment for a new internet service provider. It almost succeeded, finding a $100 promotional discount in a mass-deleted email, but got stuck at a “human check.” The bot explained, “This particular check needs a sustained mouse hold that my browser controls don’t support,” prompting the author to manually hold a button in the browser to proceed. Once past that, Dot reached the checkout screen and asked for bank account details. Unlike Muse, there is no slick Stripe integration to save credit card info and use a virtual card. Since the new provider needed a bank account to secure a monthly discount, Dot paused to let the author enter the numbers. The author decided to enter the data themselves on their own computer rather than trusting the virtual browser.
Testing Dot with a few other personal tasks yielded mixed results. It failed to find the option for a free tour and entry at a coworking space, instead offering to sign the author up for a $35 day pass. The author questioned Dot over the phone, and the bot replied, “I jumped to the wrong conclusion.” Later, the author asked Dot to retrieve information from their Ikea account about a hanging planter. It reported being unable to log in due to a looping security check. Additionally, a local teriyaki spot blocked Dot’s access to its ordering page. Something about Dot’s footprint seems to get it tripped up on security checks more often than Muse or Instinct; Dot offered a link to OpenAI’s page about websites blocking its cloud browser traffic as an explanation.
The author saw Dot come into its element when given access to something fully within their control: their personal website. The site was embarrassingly outdated, and the author asked Dot to pull together a new, streamlined design. It did, and while typing out changes for the next version, the author decided to just give it a call. Over 10 minutes, the author hashed out the desired changes section by section while doing dishes, and received a ping that the next iteration was ready for review. The results were better than manual work. The author carried out more rounds of iterations chatting with Dot in a browser window and using voice-to-text on the phone app while cleaning up.
Dot made a lot more sense once given access to the author’s computer. There, the author assigned two tasks: consolidating a handful of video files on the desktop into one clip sized for social media, and deploying changes to the personal website prototype. While the author had to grant a shocking number of permissions, the bot succeeded. It clicked around the website backend and recreated the prototype block by block. The author walked away and received updates on the ChatGPT app; the consolidated video clip was ready for review. While the author could have figured out the video editing themselves, Dot’s result was much better and upload-ready. It handled the website update as well, creating a basic design miles ahead of the previous complicated version. The author had to hand it to Dotty McDotface.
Dot makes a lot more sense when used as “Codex, but for regular people”—a cute little blob that can handle some business while you do other things. Where more personal AI agents like Muse feel like glorified autofill, Dot seems like a harbinger of the future of work. Presumably, since you pay for it upfront, it won’t be subject to the monetization pressure that Muse and Instinct will eventually face, which makes sense for enterprise software. It helped the author with tasks they might not have otherwise done, which is handy, but not $100-a-month handy in the author's line of work. It would likely be much more useful to entrepreneurs like Casey Newton who are launching new businesses. Ultimately, with some help logging into an Uber Eats account, it did actually order teriyaki.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.