The always-on AI personal assistant, which keeps working in the background even when you aren\u2019t asking it to, is this year\u2019s answer to the question of how people will actually use deep learning models. The startup Instinct became an overnight sensation, and Meta put the full force of its marketing behind its assistant, Muse.<\/p>\n
OpenAI wasn\u2019t going to be left out. The company released dots<\/a>, its assistant offering, at its developer day in late September. Alexander Embiricos, who leads OpenAI\u2019s enterprise product team, joined us at TechCrunch Disrupt this month to explain where the product came from\u2014starting with the question of why so many companies seem to be turning to cutesy mascots as avatars for their trillion-parameter matrix multiplication machines.<\/p>\n \u201cThere\u2019s kind of this convergent evolution happening,\u201d Embiricos said of the efforts to build AI assistants. \u201cAs we were working on the product, we realized if we could frame this as, it\u2019s an entity and it\u2019s a single conversation, not many different conversations\u2026.it starts to become intuitive.\u201d<\/p>\n The assistant is available in the ChatGPT desktop and mobile apps to subscribers on the Pro tier or higher, with plans starting at $100 a month. Once it is connected to your calendar, email, computer and other tools, the agent tries to help you get your work done.<\/p>\n Embiricos points out that you could achieve much of dots\u2019 functionality with Open AI\u2019s Codex app, its coding tool, released at the beginning of this year. But that requires more work and hand-holding. He recalled celebrating when his wife started leaving her computer on all night so that her agents could keep working.<\/p>\n \u201cOn the product team, what I think about is, how do we make this intuitive and useful to people?\u201d he told Disrupt attendees. \u201cA lot of that isn\u2019t building some new piece of tech into the product; it\u2019s how you frame it to the user and how it feels.\u201d<\/p>\n Making the technology feel intuitive meant repackaging the functionality of the company\u2019s models into something more accessible, a path it was already on as it spun ChatGPT Work<\/a> out of the agentic coding tools it had built for software engineers. Even the name\u2014a play on the negative space inside of OpenAI\u2019s logo\u2014is intended \u201cto feel like just something that could eventually be ubiquitous.\u201d (Asked when he learned that rival AI entrepreneur Elon Musk owns the dot.com URL, Embiricos said, \u201cI found out very late.\u201d)<\/p>\n The idea of a persistent agent is that it doesn\u2019t need prompting. For OpenAI, means more token usage, and for users, a more pro-active assistant \u201cthat\u2019s sort of thinking about you and thinking how to help you,\u201d as Embiricos put it.<\/p>\n In our testing, the assistant\u2019s ability to think ahead is apparent and often useful, but it isn\u2019t quite there yet. It will flag calendar conflicts and upcoming deadlines, but not necessarily in the most timely fashion.<\/p>\n \u201cThe first time it told me about one of those [conflicts]\u2014this is like early in development\u2014it told me three minutes before the thing,\u201d Embiricos says of his dot. \u201cIt\u2019s like, \u2018Hey, this thing is coming up. You might want to do something about it.\u2019 I only have three minutes!\u201d<\/p>\n Embiricos expects the timing to improve over time, but the proactive approach is key to the product. \u201cI was hearing about an engineer at a drone delivery startup, and they were saying, \u2018We had this outage, and by the time I noticed that the outage was going on, my dot had already noticed the outage was happening, had already analyzed the outage, and had come to me with useful information for how we could resolve the outage.\u2019\u201d<\/p>\n Giving dot the keys<\/p>\n