When Microsoft partnered up with OpenAI from a commercial standpoint in 2016 and went on to consolidate the same by investing $1 billion in 2019, there was no foreshadowing of Microsoft’s OpenAI bet becoming a historical investment.
The startup landscape had become an average corporate ventures market, hot on the heels of artificial intelligence, electric vehicles, aerospace and advanced logistics becoming cornerstones of monumental valuations.
Following the turn of events of multiple market collapses, the once-envisioned high-growth tech companies are losing money but AI seems to be sheathed in ironclad armor.
San-Francisco-based OpenAI vaulted to an overnight phenomenon which doesn’t give the impression of being transient when it introduced ChatGPT, a language model that converses with crafty human-like replies. And three years later, the taciturn partnership of OpenAI beckoned a $10 billion fat investment from Microsoft.
Now the erstwhile under-the-radar investment announcement of Microsoft, stimulated conversations in venture circles and stakeholders alike, who were trying to wrap their heads around the potential value of their stock. The Microsoft OpenAI deal’s cumulative investment has ballooned to $13 billion whilst OpenAI’s valuation has scored $29 billion.
OpenAI, Microsoft And Industry Concerns
On pace to generate $200 million in revenue in 2023 and anticipate $1 billion in 2024, OpenAI’s swift surge of 150 percent from the previous year implies growth of nearly 400 percent.
“Our plan is to be a big independent standalone company, with no plans to go public or get acquired.”
Founded as a non-profit organization in 2015, OpenAI was nowhere close to going public. But in 2019, the formation of a ‘capped-profit’ entity named OpenAI LP was making rounds, restricting the startup’s first investors from divesting more than 100 times their wealth, prompting lower returns for later investors like Microsoft.
Elon Musk, who was one of the prolific founders of OpenAI, rebuked the AI company’s unconventional structure and its implications for AI, given the level of ownership swooped under Microsoft. While maximizing returns has always been the prime concern, this unfamiliar model stirred conversations in the venture community of Silicon Valley.
OpenAI was created as an open source (which is why I named it “Open” AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, maximum-profit company effectively controlled by Microsoft.
Not what I intended at all.
— Elon Musk (@elonmusk) February 17, 2023
Previously the vice president of research at OpenAI, Anthropic cofounder Dario Amodei also expressed his concerns over the unbridled power of AI and signified Anthropic’s intentions of pursuing profits whilst having positive impacts on society.
“We will always be focused on developing innovative structures to provide incentives for safe development and deployment of AI systems.”
Implications Of The Microsoft-OpenAI Deal
As the exclusive provider for OpenAI’s research cloud computing power, products, and programming interfaces for developers, Microsoft’s bet on the artificial intelligence research laboratory is implicating a golden goose in the making. Multiple companies are rushing to incorporate their products with OpenAI, propelling massive workloads on Microsoft’s cloud server Azure.
Brockman said on Reddit that if OpenAI succeeds, it could “create orders of magnitude more value than any company has to date.” As a major OpenAI investor, Microsoft would benefit.
Microsoft’s Bing And AI
In February 2023, at a press event in its HQ in Redmond, Microsoft announced newer additions powered with AI, to its Bing search engine and Edge browser. OpenAI’s CEO Sam Altman was featured as one of the key speakers at the event.
According to an analyst at Wells Fargo, Microsoft’s move to integrate artificial intelligence into its Bing search engine, GitHub coding tools, Azure cloud and the Microsoft 365 bundle could correspond to an addition of over $30 billion in Microsoft’s annual revenue, owing half of it to Azure.
Yet, some of the highly publicized harrowing conversations with Bing’s AI failure with incorrect responses at the launch have ostracized Microsoft’s Bing towards disappointment.
A Bumpy Road For ChatGPT Rivals
Google’s rollout of ChatGPT’s rival AI chatbot Bard faced a similar fate when people hailed CEO Pichai’s decision as hasty and botched.
Another tech folly in the generative AI market was when Baidu canceled its chatbot Ernie bot’s webcast launch, dropping shares.
Other Serious Implications Of The Microsoft OpenAI Bet
Leaning on OpenAI has the potential power to spectacularly reverse Microsoft’s fortunes in AI so far, with the blocks on the road when Microsoft had to pull the Clippy assistant from Word, Twitter’s Tay chatbot and Windows’ Cortana. In October 2022, Microsoft CEO Satya Nadella confirmed that Azure Machine Learning service’s revenue had doubled consecutively for four quarters, albeit the non-disclosure of the scale of Microsoft’s AI business.
At Microsoft’s annual shareholder meeting held in December, right after the monthly frame of ChatGPT’s release, Nadella spoke animatedly of Azure’s supercomputer being utilized to train the growing qualms of the popular AI ChatGPT.
Large Language Models: ChatGPT VS GPT-3
The palpable vigor for new technologies based on LLMs, short for large language models such as ChatGPT, is rippling across the tech industry.
OpenAI’s bot, based on the core of an LLM named GPT-4, has trained extensively on online information sources to learn how to create texts that sound natural. While Microsoft has an exclusive license of GPT-4 and other OpenAI models like ChatGPT, the skeptical concerns of people describing the tech as ‘biased, deceptive, and a risk to privacy’ are slating them to call on the Federal Trade Commission (FTC) to cease OpenAI from any commercial releases of GPT-4.
In the debate between ChatGPT and GPT-3, ChatGPT looks like the obvious winner as it may be smaller than GPT-3 (20 billion against 175 parameters), but its accuracy in conversational tasks and speed are superior to GPT-3.