According to prevailing wisdom, Microsoft Corp. has emerged triumphant in the recent OpenAI controversy. However, our perspective differs significantly.
Before the upheaval involving OpenAI CEO Sam Altman and the resulting public outcry, both Microsoft and OpenAI found themselves in less than ideal circumstances. They were actively shaping the narrative in the field of AI, making substantial strides in business growth, the adoption of artificial intelligence, and operational efficiency.
Our assessment, based on conversations with clients and industry insiders, leads us to believe that this partnership has compromised its significant advantage. It is now more evident than ever that a single large language model cannot dominate the entire landscape, despite Microsoft CEO Satya Nadella’s efforts to make the best of a challenging situation.
In this in-depth analysis, we delve into the repercussions of the OpenAI debacle, exploring how it has altered customer perceptions and reshaped the competitive dynamics in the quest for AI supremacy. Additionally, the proficient data team at Enterprise Technology Research conducted a swift survey among OpenAI Microsoft customers to gauge their reactions, and we will disclose these findings.
The intricate framework of OpenAI undergoes scrutiny.
Much has been written about this rapidly evolving saga. However, it is crucial to dissect the unique and convoluted structure that OpenAI has adopted, leading to breakdowns in communication, ineffective leadership, and disastrous marketing strategies.
With the noble objective of advancing beneficial artificial general intelligence for the betterment of humanity, OpenAI was established as a non-profit entity in late 2015. Typically, government funding supports such initiatives, but due to the absence of a clear public funding route, several private industry stakeholders decided that a 501©(3) structure would be most effective. However, it became evident that the organization could not secure adequate funding to sustain its lofty ambitions, despite the initial $1 billion funding from its founders.
Microsoft, a long-standing partner of OpenAI since 2016, devised a structure in 2019 that enabled it to accelerate its external investments. The non-profit entity, which also holds a distinct holding company facilitating internal opportunities for OpenAI, expanded its framework to include an LLC with an apparent administrative role.
Familiar investment names like Andreessen Horowitz, Tiger Global, Thrive Capital, and Khosla Ventures entered the scene. Around this time, Microsoft injected approximately $1 billion into OpenAI, although it remains unclear whether this investment pertains to the intermediary entity depicted in the graphic as the holding company or the lower entity, the capped-profit company with a 49% stake in Microsoft.
A closer examination of this structure reveals that Microsoft lacks access to certain assets, such as the intellectual property associated with artificial general intelligence, which is likely segregated by the holding company. As per discussions with insiders, Microsoft’s 49% ownership of the capped-profit company is structured in tranches contingent upon OpenAI achieving specific milestones in exchange for up to $13 billion in cash and in-kind contributions, such as Azure services.
As per our understanding, Microsoft possesses the rights to derive generic applications from OpenAI’s intellectual property, potentially integrating it with Bing and other offerings, and possibly leveraging more liberal intellectual property terms. Presumably, rights are also included in the event of OpenAI’s insolvency.
Consequently, this intricate setup raises numerous concerns. Microsoft has relinquished control over OpenAI’s operations or perhaps turned a blind eye to the associated risks in exchange for exclusive access to the company’s GPT 3.5 and MPT 4 technologies. This situation enabled a board with questionable judgment to attempt the removal of the CEO, precipitating the current state of affairs.
Ilya Sutskever, a co-founder and prominent scientist, expressed regret in the following statement:
“I sincerely regret participating in the decisions of the board. My intention was never to harm OpenAI. I cherish the collaborative efforts and will exert all efforts to steer the organization back on track.”
November 20, 2023, Ilya Sutskever (@ilyasut)
A Twitter comment likened this transformation of Ilya to Barzini to Fredo, referencing “The Godfather.” While Fredo Barzini was the sympathetic yet misguided son of Don Corleone in the movie, Ilya’s fate is unlikely to mirror Fredo’s tragic demise.
The OpenAI board’s pivotal role is under scrutiny.
Let’s delve into why this debacle holds such grave implications and provide insights into why the assertion that Microsoft emerges as the clear victor may not be entirely accurate.
Microsoft essentially serves as the public face of OpenAI. The two entities are intricately intertwined. Consequently, Microsoft directly benefits from the deployment of OpenAI tools, given its exclusive partnership for the GPT technology utilized by the company.
OpenAI has achieved remarkable success, as evidenced by data from ETR’s Emerging Technology Survey encompassing over 1,500 IT decision-makers evaluating their adoption of technology solutions. ETR gauges Online Sentiment and Mind Share, represented by the blue and yellow lines, respectively. OpenAI’s inclusion in the February 2023 survey has propelled its growth on both fronts to unprecedented levels.
The black and dark bars signify no adoption, while the green and blue bars indicate incremental adoption. The upward trajectory of OpenAI’s adoption is unmistakable.
OpenAI’s dominance appeared unassailable until now.
The competitive positioning of OpenAI vis-à-vis its rivals is particularly noteworthy.
The chart above illustrates data on Net Sentiment (Y) and Mind Share (X) among leading AI players. A glance at OpenAI’s standing reveals its superiority over established and financially robust language model players like Anthropic and Cohere on both metrics.
The symbiotic relationship between OpenAI and Microsoft, bolstered by the latter’s extensive resources, software portfolio, and cloud infrastructure, reinforces OpenAI’s formidable position.
Trust in AI is paramount. Trust acts as a critical enabler.
Customers encounter significant challenges when transitioning conceptual AI into practical applications, primarily stemming from contentious legal and compliance issues: privacy concerns, security risks, regulatory compliance, accuracy limitations, unexpected outcomes, and ethical considerations. The recent PR debacle has heightened these apprehensions among customers. The hurdles to adoption are predominantly driven by these factors, as depicted by the bars in the chart below.
Following a turbulent Burn conference, Nadella has been diligently working to salvage the situation. In his conversation with Kara Swisher, he emphasized:
“We will not allow ourselves to be caught off guard like this again; I want to make that abundantly clear. That chapter is closed…”
This predicament has engendered a trust deficit for Microsoft. Consumers are urged to reflect on this crisis carefully. Can they trust Microsoft with their data and potentially entrust their company’s future technology roadmap to an entity that seemingly disregarded management risks in a multi-billion dollar investment, while OpenAI grapples with a dysfunctional board and organizational structure?
Undoubtedly, such uncertainties cast a shadow over OpenAI. The onus now rests on Microsoft to address these customer apprehensions effectively.
Winners and losers alike
Who stands to gain and who stands to lose in the aftermath of the OpenAI board’s upheaval?
Winners
Recent developments have brought a modicum of relief to stakeholders closely associated with Microsoft and OpenAI. Amazon Web Services Inc. is poised for a significant impact in 2023 at Invent: Next, Google LLC is on track to rival Vertex AI, and despite facing scrutiny from its backers, Anthropic stands to emerge unscathed if it navigates the current challenges successfully. Tesla Inc., IBM Corp. with Granite, Llama 2 descriptive channels, and other open-source initiatives…
Oracle Corp. and Salesforce Inc. are actively engaged, while Dell Technologies and Hewlett Packard Enterprise strive to implement relational AI on-premises, catering to environments where substantial data resides. Nvidia Corp. and other chip manufacturers seek a level playing field with their clientele to thrive. Visionary leaders in the artificial intelligence domain, such as SAP CEO Marc Benioff, are actively recruiting talent.
The AI community, often advocating for cautious progress, is generally content with this setback, and these entities are poised to reap benefits in the near term.
Losers
However, what about the casualties? The OpenAI board members, investors, and employees were all eyeing the lofty $90 billion valuation that now appears to be slipping away. This valuation is likely to witness a downward revision. The discourse surrounding accelerated AI development and the sector’s ability to self-regulate has been derailed, leaving users caught in the crossfire.
Despite Nadella’s agile response in the eleventh hour, Microsoft cannot be entirely pleased with the unfolding scenario. While Nadella’s strategic maneuvers are lauded as a significant triumph, it was the AI puppeteer who emerged from the Burn conference, only to grapple with the challenge of reassembling the fragmented pieces.
What if Sam Altman were enticed to return to OpenAI? Would that resolve all issues? Can operations resume without disruptions? Will the integration of settings proceed seamlessly? Should Altman align with Microsoft, how swiftly can the company scale up, and how many OpenAI personnel would transition to Microsoft? The preferable scenario—past or present—remains a subject of conjecture.
Customer-centric perspectives
Amidst the tumult, customer sentiments have largely remained unexplored.
Let’s delve into this aspect. To gauge the sentiment among OpenAI/Microsoft customers, ETR swiftly mobilized a real-time feedback mechanism. While these responses carry a subjective tone, the overarching sentiment suggests that Microsoft is deeply entrenched in many customers’ workflows, prompting several to overlook recent events. Nevertheless, concerns regarding transparency persist.
Here are excerpts from various IT decision-makers, including senior architects, seasoned IT directors, and cybersecurity specialists hailing from prominent organizations in the financial services, insurance, and philanthropic sectors.
“Three CEOs in three days signal trouble, in my opinion, and Microsoft benefits from OpenAI’s missteps.”
“Our organization places a premium on effective management, governance, and security, leading us to collaborate with Microsoft for robust solutions.”
“The lack of transparency from OpenAI regarding recent decisions casts doubt on my trust in either party.”
“The simplest recourse might be to steer clear of both OpenAI and Microsoft in favor of alternative platforms until the dust settles or clarity emerges on the sudden shifts. Such revelations sow skepticism regarding these platforms.”
“In essence, we must hedge our bets across multiple vendors. Given our heavy reliance on both Google and OpenAI, the market volatility necessitates diversification, as there is no clear frontrunner yet.”
“This underscores not only OpenAI’s governance shortcomings but also how these failings stifle innovation across the broader tech ecosystem. Policymakers advocating for AI regulation now wield greater influence.”
Monitoring the evolving narrative
What catalyzed this chain of events? Twitch’s interim CEO, Emmett Shear, initially expressed confidence in the situation upon assuming the role, citing Sam Altman’s well-being. However, he now indicates his departure from the board before comprehending the precise rationale behind his removal.
What transpired?
The value of OpenAI has been significantly eroded, potentially halving or worse. Preliminary insights suggest that some buyers are hitting the pause button, reevaluating their positions to identify potential blind spots.
For AI strategists advocating for robust AI governance, this episode serves as a poignant “we told you so” moment. While far from ideal from a technological standpoint, it severely impairs the tech sector’s capacity for self-regulation. Let’s face it: the tech industry lacks the adept lobbying prowess of financial institutions, casting a negative light on the sector.
Ultimately, there won’t be a single dominant language model to rule them all, as we have iterated throughout this narrative. In the realm of general AI, the operative models will pivot towards choice and flexibility.
Keep in touch
Next week heralds the commencement of Supercloud 5: The Battle for AI Supremacy. Live from our Palo Alto workshop, John Furrier and I will be reporting from Invent in Las Vegas alongside SiliconANGLE contributors Rob Hof, Mark Albertson, Lisa Martin, and Savannah Peterson. Rob Strechay and Rebecca Knight will provide coverage from Barcelona at HPE Discover. Tune in to the kickoff on Tuesday, November 28.
Special acknowledgments to George Gilbert and Sarbjeet Johal for their invaluable contributions this week. Gratitude to Alex Myerson and Ken Shifman for their support in managing manufacturing, podcasts, and media processes for Breaking Research. A heartfelt thank you to our Editor-in-Chief at SiliconANGLE, Rob Hof, and to Kristen Martin and Cheryl Knight for their unwavering support in keeping our team informed and amplifying our reach.
Remember to catch our weekly releases on SiliconANGLE and Wikibon. Our events are available as podcasts across various platforms.
David, Internet Commentator on our LinkedIn posts at [email protected], and DM @dvillante on Twitter.
Take a moment to explore the ETR Tutorial we crafted, delving deeper into the savings process. Kindly note that ETR operates independently from SiliconANGLE and Wikibon. For inquiries regarding identification, reproduction, or additional information about ETR’s services, please reach out to [email protected].
