Disclosure: Microsoft is a client of the author.
I remember when Office was first announced and released. Compared to the mess of applications that came before it, it was a Godsend. Yes, Lotus Symphony was on the market and many of us preferred it — at least, those of us in accounting because it was spreadsheet-based. But it would not have arrived at all had not Office shown up first.
Office turned out to be better for most users. Even with challenges from rivals like IBM and Google/Alphabet, it held on and remains the standard when it comes to a set of personal productivity tools. That’s true on Windows, macOS, iOS and Android’s different forms.
This week, Microsoft announced that Microsoft 365 Copilot is coming before the end of the year. It has the potential to do for Office what Office did to the products that preceded it — make everything else obsolete. (Though first it will need to overcome the negative response to the $30 a month per user cost, which was not presented well.)
More about the pricing issue in a minute
Microsoft 365 Copilot’s role
Microsoft has a number of ChatGPT generative AI offerings coming to market. There’s Bing Chat (a free utility provided to Bing users), Bing Chat Enterprise (free to most under Microsoft desktop licenses, except for government and education), and a number of Copilot offerings, the latest being this one.
Bing Chat Enterprise will eventually have a $5 incremental per user price for those not under one of the included licensing programs; it provides the benefits of Bing Chat, but secures internal data. The data is not shared externally or used to train the primary generative AI platform. This process has been tested in Europe and, Microsoft says, complies with GDPR and European Data Boundaries laws.
It essentially allows a user to find information, both internal and external, with the same intelligent ChatGPT engine used in Bing Chat.
Microsoft 365 extends beyond that by including Bing Chat Enterprise and adding Microsoft 365 Graph and various Microsoft 365 apps. That allows users have AI help with projects done in Office. The user and company administrator remain in control, and productivity benefits should be significant.
Early problems and limitations
The initial limitation involves an inherent inability to learn about the user so that Microsoft 365 Copilot’s output looks and feels like it was created by a person and not by a machine. Since company-specific data, including user data, stays with the company, training an AI tool on that data is problematic. So, at least for now, one of the highly anticipated features — the ability to act and appear as a proxy for the user — is not yet enabled.
The other issue is, as mentioned, price. Since generative AI is new, particularly new to users, the $30-per-user cost for Microsoft 365 Copilot seems excessive. Microsoft did not first lay the groundwork for that price by highlighting it’s less than other, more limited, alternatives in the market. In addition, Microsoft could have highlighted the productivity benefits of using generative AI to justify its price.
There is substantial value with this offering, but IT and the financial folks most IT executives report to make decisions based on hard numbers — and those numbers were not part of the price announcement.
Think of a typical TV commercial. The announcer will present a product, convince you it is worth a certain price, and then announce that the actual price is far cheaper and, if you buy now, you get even more stuff. The way to announce a price is to do so after establishing the product’s value. That was not done here, and it may slow initial sales.
With time, a game changer
That said, generative AI is a real game changer. On paper, Microsoft 365 Copilot is no exception. And while it is in late beta now, it should see general release before year’s end. Once users get comfortable with it, they should see a significant performance boost because Copilot will increasingly be able to do an ever-larger number of tasks autonomously.
Once Copilot is in broad use, expected productivity improvements of 30% to 80% become increasingly possible and could yield a boost in employee productivity not seen since Office was first deployed. The next problem will be how to manage the massive increase in documents likely to result from this productivity boost; I expect storage vendors will like this offering as much as users once they become familiar with it.
By 2030, I expect there will be little we can do without the help of generative AI. Microsoft 365 Copilot is just the start.