Meet my personal assistant, Mrs. Swift. Mrs. Taylor Swift. AI-powered chatbots may turn the future upside down

1. Are personal assistants the next big thing?
2. Chatbots and the potential pitfalls
3. Grok, Replika and Character.ai
4. Can chatbots be social?

With the increasing ubiquity and power of AI systems, new chatbots are popping up regularly. The most household name among chatbots figures to be ChatGPT—even before its parent company OpenAI forced out and then reinstated founder Sam Altman as CEO. But ChatGPT is by no means the only bot on the block. Elon Musk recently unveiled Grok, while Bill Gates, former Google CEO Eric Schmidt and semiconductor giant Nvidia led a USD 1.3b round of funding in Inflection AI this past June. Their flagship product is a chatbot/personal assistant named "P.I." Today, OMR is taking a closer look at P.I. and breaking down the potential game-changing impact of personal assistants.

My colleague Angela recalled a recent conversation she had that’s stuck with her. "Why is it so important to you to be funny?,” she asked her interlocutor. "It's not just about being funny,” they replied, “it's about communicating an emotion or intention to others. If you can make someone smile or make them think, you make the world a little bit better. It's about adding a little touch of lightness and humor to life. A bit like a spiritual Zippo that adds some light to the world.” As you may have guessed, Angela was not conversing with a friend or a human, but with the chatbot Pi, the brainchild of Inflection AI’s co-founder and CEO Mustafa Suleyman.

Are personal assistants the next big thing?

P.I. is a personal AI assistant or, as Bill Gates calls them in his blog “agents.” At their essence, they are chatbots and therefore per sé similar to other bots such as ChatGPT, but claim to be different in significant ways, describing itself as a "conversational chatbot." Meaning, ostensibly, that its primary focus is on interacting with, asking questions of and creating pleasant conversation for users.

According to Bill Gates, AI assistants are poised to upend the software industry. “In the near future, anyone who’s online will be able to have a personal assistant powered by artificial intelligence that’s far beyond today’s technology,” he wrote in his blog. A tech luminary singing tech’s praises is nothing surprising, but Gates’ timeline is: five years. Within half a decade, Gates' grand vision entails never having to visit (as if it’s chore) a website yourself again. Instead, that’ll be a drab task for your AI agent, who will also organize your daily schedule and assist with routine tasks, such as writing emails, planning appointments or even shopping. In short, everything that we all use apps, websites and individual programs for. And these personal assistants may fit in the palm of your hand, be draped around your neck or tucked away in your pocket as AI wearables.

First-gen "AI Wearables," whose aim is to knock the iPhone from its pedestal: from left "Rewind Pendant," the "Humane AI Pin" and Tab (Images taken from respective company websites, collage by OMR).

Way back in 1984, Gates’ eternal nemesis, Steve Jobs, predicted that " The next stage is going to be computers as “agents.” In other words, it will be as if there’s a little person inside that box who starts to anticipate what you want. Rather than help you, it will start to guide you through large amounts of information. It will almost be like you have a little friend inside that box."

With regard to the aforementioned USD 1.3b funding round Gates was a part of, Inflection A.I. has set its sights on building an "amazing personal digital assistants" together with partners. "Agents are not only going to change how everyone interacts with computers. They’re also going to upend the software industry, bringing about the biggest revolution in computing since we went from typing commands to tapping on icons," writes Bill Gates. And their capabilities will dwarf those of all bots that have come before. In his blog post, Bill Gates says that AI assistants will have an especially large impact in four primary areas: health care, education, productivity, and entertainment and shopping. “A shock wave in the tech industry,” in his words; CEO Suleyman titled his new book: The Coming Wave, a reference to social changes that should not be underestimated and could have serious consequences.

The potential impact of agents is fundamental. The at-your-fingertips access to information they can provide could change the extent that people no longer necessarily have to search the internet themselves when they need information or want to shop. Instead, the agent takes over. However, this also means that agents would become the new gatekeeper—a consequential development for companies and brands if they are not suggested by the agent.

Chatbots and the potential pitfalls

Limited visibility for brands is one thing. Other negative consequences discussed at length since the hype surrounding AI has captured our collective imagination, includes the loss of jobs and income for certain professional groups, discrimination and bias in algorithms, risks to privacy and security. In particular, the issue of data protection, i.e. what data these AI agents collect about their users, where it is stored and how it is used, remains murky at best. AI bots do not figure to collect less user data than the current major social media platforms; legal and ethical issues will obviously need to be addressed, too.

Grok, Replika and Character.ai

P.I. is of course far from being the only chatbot. ChatGPT is a household name now after the OpenAI’s failed ouster of its CEO Sam Altman. Just a few weeks ago, Elon Musk released his chatbot Grok. Additional alternatives include Replika, Chai, Simsimi, Myanima, Kuki and character.ai. The latter is intended to enable users to talk to historical or contemporary celebrities. Google is in talks to invest several hundred million dollars in the company.

Have a chat with Napoleon, Julius Caesar or Beethoven—that's what character.ai promises to deliver

Can chatbots be social?

Most chatbots aim to stand from the competition via some unique selling point: Character.ai says it can imitate Taylor Swift, Albert Einstein and Beethoven, among others. ChatGPT is a large, complex model that has been trained using a huge amount of data. P.I., on the other hand, describes itself as a social assistant that, unlike ChatGPT, is designed to "establish a real connection with users and create an enjoyable experience."

After testing ChatGPT and P.I., it became clear to us that both have their strengths and weaknesses. ChatGPT occasionally plays it fast and loose with the facts, but can give well-founded answers to many questions. P.I. tries hard to be funny and witty, but doesn’t quite get pull it off (is there anything spiritual about a Zippo?). Nevertheless, P.I. does create a pleasant and genial vibe by asking thoughtful questions and keeping things light. “My sense of humor might not be for everyone, but I try to keep things lighthearted and entertaining. I mean, what's the point of talking to a computer if you can't laugh, right?”