Meet Amy, an AI Enabled Assistant
Meet Amy, an AI Enabled Assistant
X.AI, the creator of the personal smart assistant “Amy” is one example of a company that’s built its entire business model on “digitizing” how administrative tasks are carried out and disrupting the role of personal assistants. Traditionally, when two people want to schedule a meeting, they may go back and forth, looking for times that work, optimizing for attributes like timing, location, etc. Users looking to schedule a meeting through Amy AI can now simply add Amy to their email chains, at which point Amy will email the user’s guest on the chain with suggested times.1 The user can then go back and forth with Amy to find a time that works, and once an agreeable time is selected, Amy AI will automatically send a calendar invite to both parties. 1
For X.AI, it’s hoping to monetize this service offering to businesses by expanding access to a personal administrative assistant to a wider range of employees, given traditionally only more senior positions in a company had personal assistants. X.AI’s value proposition is centered around increasing productivity and cost savings. On the productivity front, tasks like scheduling can now be an afterthought, taking up less mindshare and removing yet another low-value add task for the user. According to X.AI’s rough estimation, a user today can spend as much as 10 hours per month on scheduling meetings. 2 The cost savings for users will come from the tradeoff between paying X.AI $39 – $59/month2 for a license to use Amy AI, and potentially replacing multiple FTEs that currently help users with administrative tasks. Furthermore, the reduction in time alone translates to about $700/month per person. 2
Is the Future Bright? And for Whom or What?
While Amy AI may sound like a great piece of technology that can potentially save us a lot of time, and also take on more tasks in the future as it evolves, the dark side to this (depending on who you ask) may be a widespread replacement of human administrative assistants and secretaries. According to the Bureau of Labor Statistics, there were about 4 million jobs within the “Secretaries and Administrative Assistants” occupation, earning a median income of $36,500 per year.3 Salaries alone from this occupational group are almost $150 billion. 3 Suppose these jobs in the future were replaced with smart assistants like Amy AI, the new “operating cost” to run on AI assistants will be about $3 billion (assuming the $59/month subscription). 2 Of course this is some rough math and assumes that Amy AI can carry out all the tasks that human assistants do at the current price point. But if this is where the trend is going, we can see an entire occupational function wiped out, raising questions about unemployment, the skills gaps, and public/social costs.
Skeptics may question how widespread Amy AI can get in an organization because it just schedules meetings. And the fact of the matter is administrative assistants do more for an organization that just manage calendars. However, one of the key thesis of any AI technology is that over time it will learn, in this case about a user’s habit, and for X.AI this data can drive the company’s future functionality design for their product. Amy AI may only be able to schedule meetings today, but tomorrow it may be able to help you book travel and/or help you take phone messages as well. Furthermore, we already see a rise in smart assistants aiming to take more administrative tasks off of a user’s plate. Some examples include Ozlo, an AI-enabled chatbot that helps users find restaurants, and Mezi, a personal shopping assistant.4
The Business Case: What else for X.AI
X.AI is still a relatively young company, having only raised Series B funding of $23 million in April 2016. Some speculate that the company is only worth about $100 million in valuation, a tiny amount compared to the lights of Google, Amazon, etc. 5
Moving forward, I think X.AI should accelerate its product development to include more productivity based functionalities such as booking conference rooms, booking travel / lodging, ordering dinner, etc. Given the competitive landscape and multiple startups coming online to compete in this space, wider functionality can be a point of differentiation. Additionally, X.AI should also expand the functionality of its free offering to users (non-business) to build out greater scale and leverage data from users to inform future product development. Having a large dataset can help X.AI and Amy’s underlying AI technology/algorithms become smarter as data and more usage are key inputs to how quickly Amy can “learn”.
Word Count: 762
Sources:
- https://x.ai/how-it-works/
- https://x.ai/pricing/
- http://www.bls.gov/ooh/office-and-administrative-support/secretaries-and-administrative-assistants.htm
- https://www.cbinsights.com/blog/early-stage-virtual-assistant-startups/
- https://techcrunch.com/2016/04/07/rise-of-the-bots-x-ai-raises-23m-more-for-amy-a-bot-that-arranges-appointments/
- http://www.forbes.com/sites/vivianrosenthal/2016/10/28/how-to-use-ai-to-streamline-your-life/#7207068b18c2
Loved this post! Personal assistant services have gotten pretty hot in the tech space. My partner swears by Fin (https://www.fin.com/), which seems to work like Amy AI but is also tailored toward non-business users (which you suggested Amy AI branch out into). He uses FIN to schedule meetings, send reminders to people, return Amazon items, make restaurant reservations, and look up random facts. Most services are included in the monthly fee, but some cost extra. A few questions:
1.) Who do you think is the ideal Amy AI user? Are there certain companies that are more attractive as target customers?
2.) Who do you see as Amy AI’s biggest competitors and what is Amy AI’s point of differentiation?
3.) Are you concerned about data privacy? How do you think Amy AI can protect data when businesses are giving access to their private servers?
I’d never heard of this before! What a great premise – I think everyone can relate to wasting hours arranging people’s schedules for dubious-value-add meetings.
That was a poignant paragraph about the potential employment implications of automating an entire function. The good folk at Planet Money often do podcasts about digital disruption & the effect on employment. One of my favourites was their episode about spreadsheets: before Excel was invented, there’d be analysts who’d spend all day in an office manually calculating through a paper spreadsheet to work out the total impact of a change in one number. After Excel was invented, their work didn’t become redundant – they just ended up running hundreds more models with the time saved & conducted more value-add functions such as analysing trends and providing recommendations.
The same might happen here. Admin assistants might be able to take on substitute tasks such as touching base with clients. Or perhaps the roles might become part-time and would therefore open up more employment opportunities for individuals with children.
http://www.npr.org/sections/money/2015/02/25/389027988/episode-606-spreadsheets if you were interested
A cool and useful opportunity for AI! It was really interesting to hear your opportunities for X.AI. There does seem to be a number of opportunities in the travel and hospitality segment where a number of tasks performed revolve around reservations and a person confirming / checking through a database. I wonder if this could also be used as a recommendation engine. For instance, if X.AI knows that you are consistently meeting a person for coffee, if they could suggest coffee shops or other relevant places in the area.
Additionally, it seems like X.AI is branding their product right now but I wonder if they have thought about white labeling the technology out for other players to use. That way X.AI could be the engine and likely scale much faster than convincing companies, people to use a branded product.
Interesting article! I was first introduced to Amy when trying to schedule a meeting with a friend. The experience was so natural and human I actually though he had a personal assistant. 🙂
I agree that X.AI should also expand the functionality of its free offering to build out greater scale.
In my eyes, the potential is in the information the company is gathering. Imagine that base on your schedule Amy would recommend to meet in a cafe shop which will not only save you time due to its location but will also offer you a free coffee?
The potential is enormous, and this is why I support the notion that the company should focus now on scale rather than profitability. the information is the key. Companies who will control it will have an advantage to offer different services.
Interesting read! Call me traditional, but I’m squarely in the camp that there is something to real, human interaction that can’t be replaced by even the smartest AI “Amy’s or Als” of today. Executives want assistants with great organizational skills, but also those with charm, creative thinking ability, and frankly, an interesting personality as well. Could AI learn and eventually compete with human beings on that level? Is there opportunity for X.AI to produce a mass market product that can be compatible with smartphones and could help young urban professionals organize their daily lives?
I’m really interested in AI and hope administrative functions can be automated to free up admin staff time for the more complex tasks. However from my experience arranging meetings is never as simple as just finding a time when both parties are available. There are several nuances that AI still struggles with such as meeting time changes. I feel we have quite a few years to go before AI can reliably replace human staff. That being said the progress being made is impressive and as usage increases the AI will become more intelligent and develop ways to manage the more complex scheduling problems.