Microsoft’s generative AI gambit: Everything, everywhere, all at once
Microsoft was already well-placed to lead Generative AI in the Enterprise, and its rapid-fire product announcements show it’s seizing the moment.
I wanted to follow up on my last post to discuss the other half of the Microsoft-OpenAI partnership in light of Microsoft’s recent announcements.
ChatGPT shocked the consumer world with its versatility and set off alarm bells in the business world. Google, for instance, saw the threat to its $162B Search business and declared a “code red”, fearing the same disruption its Page Rank inflicted on Yahoo. This pattern is playing out across the tech world and is driving a flood of Generative AI investment ($2B in 2022 VC spending, up 425% since 2020), despite the broader downturn. Microsoft is leading that charge and shows no signs of slowing.
Microsoft’s early bet with OpenAI paid off. Microsoft clearly delivered, both in scope and depth: the successive announcements that started with Bing Chat now cover nearly every business unit. The foundation starts with the OpenAI deal – Microsoft’s cash and compute for co-development and early access. Also critical is the Microsoft Graph, go-to-market machine, and security-built-in approach.
Generative AI everywhere
Microsoft integrated Generative AI across its stack just over two months since ChatGPT launched. The speed and breadth is head-spinning: in two months, every major line of business except Games and Security added GPT. This ubiquity is only possible thanks to a focus and coordinated execution difficult in a big company. Its competitors are now on Microsoft’s clock. Salesforce announced Einstein GPT in its Clouds a day after Viva Sales’ own announcement, and Google shared its productivity plans a day before the Microsoft 365 event, sans preview. Google Bard, its Bing Chat competitor, shipped a month late and with less functionality.
Microsoft is launching generative AI capabilities at a fast pace, all with live code. For reference, OpenAI’s GPT-3 launched June 11, 2020 and ChatGPT launched Nov. 30, 2022.
Microsoft’s SaaS products share a brand – a “Copilot” that helps you start and iterate your work. The Microsoft 365 Copilot announcement reflects this broader product strategy: it’s here to help, but you’re the ultimate decider:
With Copilot, you’re always in control. You decide what to keep, modify or discard. Now, you can be more creative in Word, more analytical in Excel, more expressive in PowerPoint, more productive in Outlook and more collaborative in Teams…
With Copilot in Word, you can jump-start the creative process so you never start with a blank slate again. Copilot gives you a first draft to edit and iterate on — saving hours in writing, sourcing, and editing time. Sometimes Copilot will be right, other times usefully wrong — but it will always put you further ahead. You’re always in control as the author, driving your unique ideas forward, prompting Copilot to shorten, rewrite or give feedback.
Microsoft frames its product intent in its vision for the future of work, but I see the AI doing two core tasks that manifest across its portfolio:
Drafting: This is the low-hanging fruit for generative AI: generating content. Give it a stimulus (creative outline or e-mail to respond to) and Copilot shows you a first draft for you to iterate. The graphic below summarizes the intent here well.
Information finding: Routine tasks like getting meeting summaries, finding information, or capturing insights become much more automated. This helps users become more productive.
Generative AI helps the user focus on improvements vs. upfront edits. Source: Microsoft.
In other words, everyone gets a great intern: they remove the routine tasks so you spend more time being creative and less time on drudgery – but you still need to define the objective and do the edits.
Also interesting is what Microsoft isn’t doing – removing humans from the loop. Copilot makes content but won’t share it without human sign-off. This seems smart – Generative AI with low or no human oversight could scare its core customers, the enterprise CIOs fearful of nonfactual or harmful output.
Foundational strengths
The main driver for these rapid-fire announcements is the Microsoft-OpenAI partnership, which gave Microsoft some of the most cutting-edge Generative AI LLMs on the market. The obvious benefit is early access to high-quality AI. The other benefit is know-how: operational knowledge on how to host LLMs, support customers during deployment, and co-develop SaaS offers with top LLM experts. This lets Microsoft rapidly roll out access to cutting-edge Generative AI across the tech stack.
The great tech needs to be relevant to a user, and the Microsoft Graph deliver that context. Microsoft hosts a company’s data, so it can infer the information – the data, projects, jargon, and communication styles that make each user unique. Those signals give users personalized, business-relevant content in a way that pureplay companies could not without deep integrations and data sharing.
Microsoft is implicitly more secure, allaying another CIO fear: data leaks and security risks. The data stays in the Microsoft boundary, governed by its security and compliance capabilities. This reduces the risk inherent in moving data to third parties as well as the internal risk of confidential information leaking to the broader org.
Microsoft’s final strength is its go-to-market. CIOs are under pressure to have a “Generative AI strategy”, yet many orgs lack the technical ability to implement it (66% of respondents in a 2023 Salesforce survey). Microsoft has the sales force and existing relationships to reach these customers, but it also has an unrivaled partner ecosystem that adapts and deploys software to meet a customer’s unique situation.
The path forward
Enterprise SaaS is usually sticky, but Generative AI could upend that dynamic as companies race to adapt to the new state of tech. The recent flux in Search caused by Bing Chat is a harbinger of that change. Microsoft is leading because of its early embrace of Generative AI and even stands to gain ground in areas it trails (namely Search and CRM). It must now deliver on the hype it generated.
Another rate-limiting factor is how much change customers can bear. What does it mean for everyone to have an intern, and how do new workers upskill? How do we digest the looming content glut, and can we trust the system security? Customers will be reluctant to adopt products that upend work processes until they can handle the disruption. Users and IT orgs need to adapt before they will adopt more disruptive products. Marketers, customer success teams, and consultants have their work cut out.
Disclaimer: At time of writing, I’m a current LinkedIn and former Microsoft employee. This article is based purely on public information. I don’t speculate on future products and business models in this post to avoid conflicts of interest.