ChatGPT: Hype or Future?

"Soon, you will be able to have helpful assistants that talk to you, answer questions and give advice. Eventually, you can have something that goes off and discovers new knowledge for you" - Sam Altman, founder of OpenAI. A scene of Star Trek pops into mind “Computer, run simulation of Klingon battle tactics.”. While nerds like my boss get excited about it, MNCs are riled up too. Microsoft has just confirmed a $10BN investment into OpenAI (Financial Times, 2023) and its new toy ChatGPT, much to the envy of the tech world. So what makes ChatGPT so exciting?

Unless you've been on a digital detox the last two months or have been living under a rock, you would have heard of the newest technology that's recently burst onto the scene and amassed millions of users within 2 months of launching: ChatGPT. Published by Tech startup OpenAI, ChatGPT is an AI-powered chatbot replicating human conversation. What sets it apart is its GPT3.5 language technology, trained by an enormous amount of data (45TB and 175BN parameters, and counting), collated from various sources across the internet. ChatGPT helps users with a wide range of tasks, from providing a simple historical fact to correcting errors in complex Python coding. Launched on November 30th 2022, the chatbot garnered over 1 million users within the first five days post-launch (Yahoo, 2022). Users flocked to test its impressive capabilities (including essay-writing college students, myself included!). Such a vast and quick adoption has stirred many to question the long-term impact of such technology, some are concerned that workers in the creative industry, such as journalists or programmers, may soon be replaced. Personally, I would be more concerned about those of us who do not practice creativity in our daily lives but that's a conversation for another time.

So where does this amazing and potentially disrupting ChatGPT come from? OpenAI was the baby of co-founders Elon Musk and Sam Altman, founded in 2015 to develop friendly AI to elevate humanity. The company received more than $1BN in venture capital from an all-star set of investors, including Peter Theil, Reid Hoffman, and Amazon Web Services. Musk abruptly left the company in 2018, leaving former Reddit CEO Altman taking sole charge ever since (The Economic Times, 2023). Ever since the seemingly immediate success of ChatGPT, many investors’ attention has been drawn to OpenAI and the AI industry in general, with many seeking to hop on the bandwagon.

Microsoft saw the potential in OpenAI back in 2019, a potential that saw them invest a reported $3BN, whilst concurrently working on a joint venture to develop a software platform within Microsoft's Azure AI technologies (OpenAI, 2019). This began an exclusive relationship that went firmly under the radar until now. The exclusive relationship has resulted in the channelling of further funding through a 'multi-year investment' for the continued development of ChatGPT software. Exact financial details have yet to be disclosed, but the deal is estimated to involve a $10BN investment at a valuation of $29BN (Financial Times, 2023) making it Microsoft's biggest AI bet yet. The extra capital will enable OpenAI to expand rapidly and further develop its product lines to begin monetising the chatbot. It may not seem obvious immediately, but this move's impact on multiple industries will likely have significant global players such as Google shaking in fear.

Microsoft hosts a user base of around 1.4BN people and businesses using various Microsoft Office Package (Teams, Word, PowerPoint, Excel etc.) components daily (Tech Advisor, 2022). Should (or rather when) Microsoft integrate ChatGPT AI technology into these programs, the way in which one crafts detailed emails, builds multi-tab financial models, or constructs an extensive PowerPoint presentation will be revolutionised, improving the daily workflows of literally more than 1BN people. Our generation will hold the front-row seats in witnessing how AI may improve our productivity, efficiency, and economic gains. The little-used 'Bing' search engine, always creeping in the shadow of its bigger, stronger, and more attractive older brother Google, is being developed to include ChatGPT functions natively (The Guardian, 2023). This comes as a direct act of war against Google, which currently derives around 57% of its $69BN quarterly revenue from Search functions alone (CNBC, 2021). It is worth highlighting it costs around 7x more per ChatGPT search query than it does Google, providing a short-term shield that shouldn't harm their bottom-line profit too harshly (Investing.com, 2023). But a successful collaboration between Bing and ChatGPT could easily pull 5-10% of Google’s reported 84% market share (Statista, 2023) very quickly, and it will only get bigger, disrupting the monopoly Google has in search.

Morgan Stanley recently weighed in on the argument suggesting large language models may pose a more medium to long-term threat to Google’s profitability. A rapid increase in AI adoption and interest may cause a knee-jerk reaction from Google to publish its own natural language models leading to significantly larger CAPEX requirements, in turn lowering operating margins and overall efficiency. Building on the earlier reference to a 7x greater cost per search query using ChatGPT, Morgan Stanley cites the reasoning to be the compute intensity of the model that must store, recall, analyse and compile data and text into a natural language format for the user to understand, with the average cost per query ranging anywhere between $0.004 and $0.044 (Investing.com, 2023). Nevertheless, despite increasing interest and the growth of ChatGPT, Morgan Stanley believes even an OpenAI search market share of 10% by 2025 would reduce Google's operating margins by a median estimate of 1.5%, news Google would likely welcome (Investing.com, 2023).

Google isn't backing down without a fight. In true form to any war between titans, Google threw a strong counterpunch to Microsoft by recently announcing their intention to release a Chat GPT competitor developed by DeepMind, which Google acquired for $500MN back in 2014. (The Guardian, 2014). This intention recently came to fruition </