TRUESHARES

ChatGPT’s Impact on Google and the Search Market

ChatGPT launched late last year to fanfare and millions of users who have since employed this large language model (LLM) to answer their many questions. Instead of typing a question into a search engine and receiving seemingly endless pages of links (and ads) ChatGPT produces one answer in a succinct and understandable manner. In this way, ChatGPT has the potential to replace traditional search engines. With roughly 84% of the share of the search market1 (as of December 2022), Google is likely concerned as it sees its primary revenue model under threat.

The greatest threat ChatGPT poses to Google is ad revenue. Because chatGPT provides one answer, it is far less conducive to digital ads, which accounted for 80% of Google’s revenue2 last year. Google has also been losing search market share1 to Bing (Microsoft’s search engine) since 2019. Microsoft owns 49% of OpenAI3, the company that created ChatGPT. To survive it seems Google will need to create its own chatbot. 

Because of ChatGPT’s inability to distinguish fact from fiction and its propensity to provide biased answers2 against women and people of color, it is not only imperfect, but dangerous. This is likely why Google has been hesitant2 to release its own version of ChatGPT — until now. Google launched its own LLM called “Bard” recently… with a factual error. This error4 lost Alphabet, Google’s parent company, $100 billion in market value and appears to have boosted Microsoft shares by 3%.

Any time a groundbreaking, headline-making technology is closely linked to a brand, that company’s value largely hinges on the success of that one new product. This association can lead to highly volatile and risky returns for shareholders, as is playing out right now with Google and Microsoft. For investors interested in chatbots without the downside risk of weighing too heavily on individual companies or products, consider looking to the supporting technologies they all have in common instead.

Running an AI-based LLM like ChatGPT and Bard requires high-performance computing (HPC), data storage, and processing power. ChatGPT uses Microsoft’s Azure cloud computing service while Bard uses Google Cloud. But underneath it all, regardless of the product, brand name, or cloud service, they all use HPC systems and data storage6 made by other companies. HPC systems, for example, require specialized hardware called graphics processing units (GPUs). ChatGPT currently uses roughly 10,000 GPUs from Nvidia5 for its AI training. Similarly, Google Cloud requires cyber security, which is powered by AMD processors7

As companies like Google and Microsoft duke it out for market share, they will need to scale everything, from data centers and processors, to storage and security. Plus, the cost of running AI chatbots5 far exceeds the cost of training them and are much higher than the cost of traditional searches. Most of this added cost, multiplied by the need for scale, goes to backend computing companies like Nvidia, AMD, and others that specialize in data, like Snowflake. One article estimates the computing industry could capture $30 billion of Google’s profit5 if ChatGPT replaced all searches.

Neither Nvidia, AMD, or the handful of other companies that create the necessary components for running chatbots feel the weight of the individual successes and failures of those end products. But, they could potentially reap the rewards of scaling the supplemental costs those companies need in order to compete. These companies and others like them make up a majority of the holdings in our Technology, AI, and Deep Learning ETF (LRNZ)*. They also allow investors to enter this nascent industry with a more widely diversified asset class and lower concentrated risk with upside potential as companies join the latest tech race.

* Visit www.www.true-shares.com/lrnz for a full list of holdings.

Important Risk Information:

The TrueShares AI & Deep Learning ETF (AI ETF) is subject to the following risks: Artificial Intelligence, Machine Learning and Deep Learning Investment Risk  – the extent of such technologies’ versatility has not yet been fully explored. There is no guarantee that these products or services will be successful and the securities of such companies, especially smaller, start-up companies, are typically more volatile than those of companies that do not rely heavily on technology. Foreign Securities Risk -The Fund invests in foreign securities which involves certain risks such as currency volatility, political and social instability and reduced market liquidity. Growth Investing Risk  – The risk of investing in growth stocks that may be more volatile than other stocks because they are more sensitive to investor perceptions of the issuing company’s growth potential. IPO Risk  – The Fund may invest in companies that have recently completed an initial public offering that are unseasoned equities lacking a trading history, a track record of reporting to investors, and widely available research coverage. IPOs are thus often subject to extreme price volatility and speculative trading.  Non-Diversification Risk  – The Fund is non-diversified which means it may be invested in a limited number of issuers and susceptible to any economic, political and regulatory events than a more diversified fund.

Sources:

  1. https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/
  2. https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-google-search.html
  3. https://www.searchenginejournal.com/microsoft-reportedly-planning-a-10-billion-investment-in-openai/
  4. https://www.npr.org/2023/02/09/1155650909/google-chatbot–error-bard-shares
  5. https://www.forbes.com/sites/johnkoetsier/2023/02/10/chatgpt-burns-millions-every-day-can-computer-scientists-make-ai-one-million-times-more-efficient/?sh=393f6aa36944
  6. https://dgtlinfra.com/chatgpt-openai-azure-cloud/
  7. https://www.amd.com/en/solutions/google-cloud-confidential-computing