The Benefits of Microsoft including OpenAI to its future products

Abstract

Microsoft announced today that it will deepen its partnership with OpenAI, the firm behind artificial intelligence systems like ChatGPT, DALL-E 2, and GPT-3 that produce art and text, by investing "several billions of dollars" over the course of "many years." According to OpenAI, it would use the freshly acquired funding to carry out independent research and create "powerful, useful, and safe" artificial intelligence. The precise sum of the contribution has not been made known to the general public. Microsoft's public relations situation isn't great considering that a week ago the company disclosed plans to fire 10,000 workers as part of a bigger cost-cutting initiative. Initiatives to cut costs are efforts to lower overall spending in this setting. What is meant to be referred to as "layoff plans" in this context are strategies to reduce overall expenditures. On the other hand, the corporation informed them of it earlier this month. In a recent interview with The Wall Street Journal, Microsoft CEO Satya Nadella talked about the company's aspirations to make the core components of OpenAI usable as commercial platforms. Any organisation in any industry will be able to build on top of them thanks to this. The corporation informed everyone in advance of this. Despite having a lower probability of turning a profit as a result of the new investment agreement with Microsoft, OpenAI will carry on as a business. With the caveat that there is a possibility that this restriction will be lowered at some point in the future, the theory suggests that backers' prospective returns are limited to no more than one hundred times the amount of money they first committed.

Introduction:

In order to speed up the artificial intelligence research being done by OpenAI and integrate OpenAI's AI systems with its products, Microsoft announced in a blog post that it will increase its investments in the deployment of specialized supercomputing systems. At the same time, Microsoft will "introduce new categories of digital experiences." Microsoft's Azure cloud platform will continue to be OpenAI's exclusive cloud provider, powering the workloads of OpenAI, which include research, products, and API services. Workloads for OpenAI will include research, merchandise, and API services. Around the world, Microsoft is a leader in the information technology sector. It was initially believed that Microsoft was working on integrating OpenAI's language AI technology into Word, PowerPoint, and Outlook as well as integrating ChatGPT with the Bing search results. This, however, didn't turn out to be the case. This theory was based on rumors that had previously been spread. Copilot, an artificial intelligence system that writes code, was developed and launched by Microsoft, which now controls GitHub. GitHub worked with us to complete this project. Additionally, Microsoft has used OpenAI innovations like GPT-3 and DALL-E 2 into a number of its own products and services, such as Power Apps, Microsoft Edge, and the soon-to-be-released Designer (Jayaramireddy et al., 2021, 8).

These projects are a continuation of the close collaboration between Microsoft and OpenAI that has existed for many years. In order to jointly develop new technologies for the Azure platform and "further extend" OpenAI's large-scale AI capabilities, Microsoft declared in 2019 that it will spend $1 billion in OpenAI (about half of which would be in the form of Azure credits). The majority of this investment would be made up of Azure credits. OpenAI agreed to provide Microsoft permission to utilize a piece of its intellectual property in exchange for Microsoft eventually selling those permissions to partners. Additionally, OpenAI agreed to let Azure train and run AI models while the business focused on creating the next-generation of computer hardware.

Microsoft said that it has built a supercomputer that was co-designed with OpenAI and hosted on Azure a year later. About a year after the original development, the announcement was made. This specific computer was among the most potent ones available anywhere in the world at the time. The Azure OpenAI Service was initially made available by Microsoft in 2021. With the help of this solution, organizations will have access to OpenAI's artificial intelligence (AI) systems, such as GPT-3, as well as other features designed with business requirements in mind, like security, compliance, and governance (Cihon et al., 15).

Benefits of Microsoft including OpenAI to its future products:

According to a New York Times report, Microsoft will have increased the amount it is investing in OpenAI by two billion dollars between 2019 and the beginning of 2023. A venture and technology incubator program focused on AI, OpenAI's Startup Fund, is another initiative that the enormous information technology business has become a major investor in. In a statement describing the decision by Microsoft to work with OpenAI, Nadella said, "We formed our collaboration around a shared objective to responsibly promote cutting-edge AI research and democratize AI as a new technology platform." Developers and organizations from a wide range of industries will have access to the most cutting-edge artificial intelligence infrastructure, models, and toolchain through Azure to build and run their apps, according to Microsoft as they enter the next chapter of their cooperation. This will enable us to realize the full potential of our collaboration.

In a news release, OpenAI CEO Sam Altman was reported as saying that the three-year relationship between the two businesses had been very successful. Microsoft shares our goals, and we are thrilled to continue our independent research and work toward creating ground-breaking AI that benefits everyone, the company said in a statement. It was previously reported that Microsoft was considering buying a 49% stake in OpenAI, which at the time would have valued the company at over $29 billion. sources claimed to have this knowledge. According to the terms of a proposal that Semafor presented, Microsoft would be entitled to receive three-quarters of OpenAI's earnings until it recovers an investment that might amount to as much as $10 billion. Semafor suggested various ideas, including this one. Other investors, like Khosla Ventures, would purchase 49% of the company's shares, while OpenAI would keep the remaining 2% for itself (Mahmoud et al., 177).

There is a lot of pressure on OpenAI to make money from the selling of its products, including ChatGPT. The company expects to earn $200 million through licensing and premium products like ChatGPT Professional in 2023, a pittance compared to the billions of dollars that have been invested in the company up to this time. This sum is insignificant. Several legal authorities have claimed that using Copilot could put businesses at risk if they unintentionally included copyrighted advice from the tool into their production software. A lot of different legal authorities have made this claim. Because Copilot removes the license from the code, it is impossible to know which code can be deployed and which might have terms of use that conflict with other code, as Elaine Atwell notes in an article that was posted on the Kolide corporate blog. This is due to the fact that it is impossible to tell which code has license removed.

Artificial intelligence-generated content has been disabled on services like Getty Images because of worry that it would have legal repercussions.

The text-generation system created by OpenAI has come under fire for its propensity to provide answers that, despite initially seeming persuasive, are not grounded in reality. In addition to concerns about the technology's legality, this objection is also voiced. Earlier this month, users were briefly barred from uploading information that was generated by ChatGPT on the website Stack Overflow, which enables users to publish questions and answers. The website said that the AI made it too simple for individuals to produce comments, leading to a deluge of responses on the website that were susceptible to interpretation. Additionally, a number of public school districts as well as at least one academic conference have deemed the usage of ChatGPT to be improper (Reed et al., 22).

Artificially intelligent generative art, like DALL-E 2, which was discovered to copy and paste data from the datasets on which it was taught, is also affected by the lawsuit. To put it another way, DALL-E 2 copied and pasted information from the datasets that it was trained on (i.e., images). Recently, a number of OpenAI's rivals were the target of a lawsuit that claimed they had violated the rights of "millions of artists" by using billions of photos from the internet to train AI art-generating tools "without the authorization of the actual artists." The defendants are being sued on the grounds that they violated the rights of "millions of artists." In this complaint, it is claimed that numerous rivals used the aforementioned images as training material for various tools, infringing on the intellectual property rights of "millions of artists."

This situation is due to the enormous expenditures involved in the design, construction, and maintenance of such large AI systems. The operational costs of ChatGPT alone, according to Altman, are "eye-watering," coming to only a few cents in compute rates for each chat. This is true even if ChatGPT handles thousands of talks every day. (ChatGPT has more over one million users as of the beginning of December.) The launch of GPT-3 has reportedly cost the company millions of dollars while this is happening. Significant legal obstacles that could potentially impede the advancement of OpenAI's and Microsoft's artificial intelligence (AI) research exist for both companies. If the proper actions are not taken, these challenges could be stalled. The United States Patent and Trademark Office (USPTO) has ruled that a work must have been created by a human being at some time in its history in order for it to qualify for intellectual property protection. This led to the removal of the copyright protection that had been applied to an AI-generated comic. And a class action complaint against Microsoft, GitHub, and OpenAI alleges that they broke copyright rules by allowing Copilot, the aforementioned code-generating system, to reuse parts of licensed code without giving proper attribution. The lawsuit alleges that by allowing Copilot to do this, these businesses broke the law. Allegations in the complaint claim that by allowing Copilot to behave in such a way, these corporations violated the law (Luitse et al., 2).

The deal's importance to the industry

The partnership between a well-known artificial intelligence research institute and a Silicon Valley tech behemoth has had a substantial impact on the rate of innovation coming out of OpenAI. This collaboration significantly impacted OpenAI. The partnership's primary goal when it was created was not to become a source of novel and cutting-edge algorithmic solutions, but this has changed over time. The collaboration was first created with the goal of combining the talents that each party possessed. If it weren't for the Microsoft-provided cloud resources, OpenAI's continued research would not be possible today. These cloud resources have been used for a variety of projects, including DALL-E, GPT, and Codex. This served as one of the main justifications for Microsoft's decision to invest $1 billion in OpenAI. The development of a supercomputing cluster that might serve as a training ground for the upcoming generation of experts in artificial intelligence was another factor. This investment allowed Microsoft to acquire a special license for the GPT-3 API, an LLM termed "Codex" that serves as the basis for GitHub Copilot, as well as the technical know-how required to integrate OpenAI's products with their own. The possibility of incorporating ChatGPT and DALL-E into Microsoft's for-profit software suite, Office 365, has been rumored. This rumor and hearsay served as the foundation for the supposition (Bhushan et al., 257). The following is what Microsoft CEO Satya Nadella had to say on the potential development of generative AI in the future:

The vast bulk of data will be produced in the not too distant future by generative models. On the other hand, because it has the ability to significantly affect our future, we must also consider how it can help us improve what we are doing right now because it has the power to change things. On the other hand, OpenAI received a $1 billion investment from Microsoft as well as authorization to use a sizable number of cloud resources that were made available by that company. OpenAI has previously said that they prefer to provide API access to both developers and businesses rather than actively commercialize their goods. Because OpenAI prefers to promote innovation above competition, this choice was chosen. OpenAI came at this conclusion after deciding that they believe this method will produce superior results. As a result of the arrangement, it is not concerned with the costs associated with cloud computing, which are necessary in order to initially train the models. This is one of their main revenue streams. Sam Altman has already stated that the compute expenses for ChatGPT are described as being "eye-watering," which is public knowledge. The Redmond-based technology behemoth Microsoft, on the other hand, reportedly intends to safeguard its bet on AI by investing $10 billion in the AI think tank over the next four years in exchange for a 49% stake in the business. Many individuals are surprised by this information. This would imply that Microsoft is supporting their AI bet wholeheartedly (McStay, 7).

Discussion and analysis:

Because of the significant impact it had on the fields of artificial intelligence and machine learning, the relationship between Microsoft and OpenAI dominated media discussion for the majority of the previous year. Microsoft has been making the most of its $1 billion investment in OpenAI, as shown by the creation of GitHub Copilot, the purchase of an exclusive license to GPT-3, and rumors regarding the integration of ChatGPT into Bing. An open-source program called GitHub Copilot aids in automating the management of GitHub repositories. On the other side, in addition to that, they have been footing the tab for OpenAI's astronomical cloud service fees because they are doing research into the most cutting-edge algorithmic advancements. The agreement has contributed significantly to the advancements in widely used deep learning since it was struck in the middle of 2019. Both businesses collaborated extensively to build the supercomputing cluster that is housed on Azure, and a sizable portion of OpenAI's capabilities have already been included into the Microsoft service offering. It has been revealed that Microsoft plans to invest an additional $10 billion in OpenAI. This was announced shortly after the corporation revealed a share offer that valued the business at $29 billion. These details have been covered in recent news publications. Given that the collaboration has achieved many of the key goals it was initially intended to, it would seem that Microsoft is eager to benefit from the value addition that OpenAI can provide to its current products and services. Regarding the current $10 billion offer, everyone is thinking the same thing: Is Microsoft planning to buy OpenAI?

What should be proven to justify a merger or acquisition?

Before we can look into the added value that OpenAI brings to the goods and services that Microsoft offers, we first need to take a closer look at the specifics of the tender offer that OpenAI has made. According to a number of industry insiders, Silicon Valley's propensity to artificially inflate the worth of its startups is further demonstrated by the $29 billion valuation of OpenAI. With the exception of the above indicated access to the API, OpenAI does not currently have a robust business model in place. Altman has stated that he will look into monetizing ChatGPT at some time in the future, although OpenAI has always had a hostile attitude toward initiatives that aim to generate profit. He has stated that he will eventually think about monetizing ChatGPT.

The business first functioned on the assumption that it would not make a profit, but then changed to a business model that only made tiny profits. The corporation's board of directors will impose a limit on profits that exceed a return of one hundred times the original investment because it is a non-profit organisation. According to reports, this action was taken in an effort to secure financing for the high computing costs associated with the creation and instruction of artificial intelligence (AI). On the other hand, this significant cost sink is no longer a problem as a direct result of the agreement achieved with Microsoft (Rikap et al., 66).

It is likely that the actual value of OpenAI's products may not justify the company's current valuation of $29 billion, depending on how the company's business fundamentals develop. It's also possible that this pricing is reasonable, though. Microsoft's consideration of increasing its investment in OpenAI and ongoing discussions about doing so show that the company agrees wholeheartedly with this evaluation. Microsoft reportedly wants to spend $10 billion in the business in exchange for a 49% ownership position and 75% of any future profits. Sources with prior knowledge of the situation have provided this information. Through this purchase, Microsoft will not only be able to acquire a sizable stake in OpenAI and a percentage of the company's future revenues, but it will also be able to expand on the advantages that its prior partnership with OpenAI has already provided. It will turn out that both parties will win in this case, which is fantastic news for Microsoft. The cascading effects that OpenAI's technologies can have on Microsoft's product portfolio, which has been lagging behind in recent years due to Google's strong competition, will also benefit from these developments. A business called OpenAI creates artificial intelligence technologies.

Conclusion:

Users of Microsoft PowerPoint will be able to include DALL-E photos into their slides, and it's possible that Adobe will lose some market share as a result of the recently announced Microsoft Designer product. Microsoft Word may finally catch up to Google Docs in terms of the autocomplete features it offers, and users of Microsoft PowerPoint will soon be able to use photographs created by DALL-E in their presentations. It's likely that Bing may once again play a significant role in the search engine business if the logical responses provided by ChatGPT and the natural language processing capabilities of OpenAI are put to use. It is very clear that Microsoft made a wise investment when they paid $10 billion for the company when one considers the algorithms that OpenAI is presently using as well as those that will be created in the future. Customers on Azure have access to extra capabilities that can be added to the application programming interfaces (APIs) that OpenAI provides. If this is finished, Microsoft might even permit business users to train their own models and make customized tweaks. This will not only boost the value added that the various Azure AI products already offer, but it will also safeguard the data security and privacy of businesses that take good care to secure their intellectual property. By creating internal think tanks or buying ones that currently exist with this purpose in mind, IT companies have already started to conduct AI research. Microsoft has been glaringly absent from the forefront of important artificial intelligence (AI) research, despite the fact that Google already has DeepMind and Meta is placing their bets on their AI lab. Despite the fact that Google already owns DeepMind, this continues. Microsoft appears to be making the right moves, levelling the playing field for both the huge technology company and the organisation that is struggling with a lack of financing for OpenAI.

References

Bhushan, Sudhanshu. "The impact of artificial intelligence and machine learning on the global economy and its implications for the hospitality sector in India." Worldwide Hospitality and Tourism Themes 13.2 (2021): 252-259.

Cihon, Peter, Jonas Schuett, and Seth D. Baum. "Corporate governance of artificial intelligence in the public interest." Information 12.7 (2021): 275.

Jayaramireddy, Charitha Sree, et al. "A Survey of Reinforcement Learning Toolkits for Gaming: Applications, Challenges and Trends." Proceedings of the Future Technologies Conference (FTC) 2022, Volume 1. Cham: Springer International Publishing, 2022.

Luitse, Dieuwertje, and Wiebke Denkena. "The great transformer: Examining the role of large language models in the political economy of AI." Big Data & Society 8.2 (2021): 20539517211047734.

Mahmoud, Ali B., Shehnaz Tehseen, and Leonora Fuxman. "The dark side of artificial intelligence in retail innovation." Retail Futures. Emerald Publishing Limited, 2020. 165-180.

McStay, Andrew. "Replika in the Metaverse: the moral problem with empathy in ‘It from Bit’." AI and Ethics (2022): 1-13.

Reed, Daniel, Dennis Gannon, and Jack Dongarra. "Reinventing High Performance Computing: Challenges and Opportunities." arXiv preprint arXiv:2203.02544 (2022).

Rikap, Cecilia, et al. "Tech giants and artificial intelligence as a technological innovation system." The Digital Innovation Race: Conceptualizing the Emerging New World Order (2021): 65-90.