Generative AI — a Review
This report will discuss a brief history of the Generative AI revolution, saturation of the space, and increasing competition. The following article will explore new, unsaturated opportunities in the space.
Let's start on November 30th, 2022. OpenAI launched ChatGPT, a more human-friendly successor to their InstructGPT model (OpenAI 2022). It enhanced conversational structure through follow-up questions and built-in "safety" features while retaining the original utility of providing detailed instructions for tasks given by a prompt (OpenAI 2022, sec. "Iterative deployment"). ChatGPT grew to over 100 million monthly active users in January, two months after the launch (Hu 2023, para. 1), and 100 million weekly active users at their DevDay developer conference in November 2023 (TechCrunch 2023).
The race for the greatest conversational Large Language Models (LLMs) ensued. Alphabet-backed Anthropic launched Claude in March of 2023, a direct competitor to OpenAI's ChatGPT, emphasizing preventing generating perceived offensive or dangerous content, such as instructions for making weapons (Nellis 2023, para. 1, 4). Inflection AI followed in May, launching Pi ("Personal intelligence"), their ChatGPT competitor, with personal and emotionally driven responses contrasting concise instruction-driven ChatGPT (Konrad 2023, para. 4). Funding followed, with Anthropic raising over $2 billion from Google and other sources (Field 2023) as well as Inflection AI closing a $1.3 billion funding round led by Microsoft, Reid Hoffman, Eric Schmidt, and others (Wiggers 2023, para. 2). These investments hedge Big Tech's own Generative AI products, such as Google's Gemini (Pichai 2023), Microsoft's Bing Chat (Mehdi 2023), and Elon's Grok (Ghaffary 2023).
Why would Big Tech pour billions of dollars into AI startups that compete with their offerings?
Mob ties. Alphabet (Google's parent company | NASDAQ: GOOGL), Microsoft (NASDAQ: MSFT), and Amazon (NASDAQ: AMZN) are publicly traded companies under intense scrutiny from regulating bodies such as the Federal Trade Commission (FTC) and the Internal Revenue Service (IRS), as well as their shareholders (CFI Team 2024, para. 8).
The stock market, where the shares of public companies are traded, is known to be influenced by "[political], social, environmental, economic, and public health factors," and public companies' stock prices are at the mercy of public opinion (Das et al. 2022, 1). This intense scrutiny makes disruptive moves dangerous.
For example, in early February 2023, just over two months after the ChatGPT release, Google teased Bard, a Google-search-style conversational AI service, to rival ChatGPT. Google shares dropped $100 Billion after Bard falsely claimed that exoplanets were discovered by the James Webb Telescope when NASA confirmed that they were discovered by the European Southern Observatory's Very Large Telescope (Olson 2023, para. 1, 6).
However, being risk-averse isn't required for public companies; it's a widely accepted style. For example, Elon Musk's x.AI released Grock, a 33 billion parameter conversational LLM for X (formerly Twitter) close to the capabilities of LLaMA 2 (70B parameters) while using "only half of its training resources" (xAI 2023, sec. "The journey to Grok-1"). This model had safeguards removed as part of his push to protect free speech; he released this model to X Premium users in December of 2023 (Perez, 2023). Musk made these risky moves while CEO of Tesla (NASDAQ: TSLA), a publicly traded company. Although Musk is a controversial figure, there are many lessons to learn from his complex relationship with risk.
Hidden truths. Peter Thiel, a famed investor and serial entrepreneur, wrote that there are two kinds of secrets, "secrets of nature and secrets about people" (Thiel 2014, pg. 103). His people's secret, he asserted, was that competition and capitalism are opposites; successful businesses are monopolies in their respect (Thiel 2014, pg. 24).
This Generative AI volume aims to map this space, providing primary sources on topics from Video Generation to Quantum AI from interviews with Professors, Executives, Researchers, and Engineers. The mission of this volume is to provide multidisciplinary clarity on the Generative AI space and promote a collective opinion backed by substance and rid of sensation.
References
OpenAI. "Introducing ChatGPT." OpenAI, November 30, 2022. https://openai.com/blog/chatgpt.
Hu, Krystal. “ChatGPT Sets Record for Fastest-Growing User Base - Analyst Note.” Reuters, February 2nd, 2023. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/.
Malik, Aisha. “OpenAI’s ChatGPT Now Has 100 Million Weekly Active Users.” TechCrunch, November 6th, 2023. https://techcrunch.com/2023/11/06/openais-chatgpt-now-has-100-million-weekly-active-users/.
Nellis, Stephen. “Alphabet-Backed Anthropic Releases OpenAI Competitor Named Claude.” Reuters, March 14th, 2023. https://www.reuters.com/technology/alphabet-backed-anthropic-releases-openai-competitor-named-claude-2023-03-14/.
Konrad, Alex. “Inflection AI, Startup From Ex-DeepMind Leaders, Launches Pi — A Chattier Chatbot.” Forbes, May 2nd, 2023. https://www.forbes.com/sites/alexkonrad/2023/05/02/inflection-ai-ex-deepmind-launches-pi-chatbot/?sh=3b5145003d6d.
Field, Hayden. “Google Commits to Invest $2 Billion in OpenAI Competitor Anthropic.” CNBC, October 27th, 2023. https://www.cnbc.com/2023/10/27/google-commits-to-invest-2-billion-in-openai-competitor-anthropic.html.
Wiggers, Kyle. “Inflection Lands $1.3B Investment to Build More ‘Personal’ AI.” TechCrunch, June 29th, 2023.
Bisson, Simon. “Inside Phi 2: Microsoft’s Small Language Model.” InfoWorld, January 11th, 2024. https://www.infoworld.com/article/3712200/inside-phi-2-microsoft-s-small-language-model.html.
Pichai, Sundar. “Introducing Gemini: Our Largest and Most Capable AI Model.” Google, December 6th, 2023. https://blog.google/technology/ai/google-gemini-ai/#sundar-note.
Mehdi, Yusuf. “Reinventing Search with a New AI-Powered Microsoft Bing and Edge, Your Copilot for the Web.” The Official Microsoft Blog, February 7th, 2023. https://blogs.microsoft.com/blog/2023/02/07/reinventing-search-with-a-new-ai-powered-microsoft-bing-and-edge-your-copilot-for-the-web/.
Ghaffary, Shirin. “Elon Musk’s Grok Represents a Serious Threat to ChatGPT.” Bloomberg, December 14th, 2023. https://www.bloomberg.com/news/newsletters/2023-12-14/elon-musk-s-grok-represents-a-serious-threat-to-chatgpt.
CFI Team. “Public Companies.” Corporate Finance Institute, 2024. https://corporatefinanceinstitute.com/resources/accounting/public-companies.
Das, Nabanita, Bikash Sadhukhan, Tanusree Chatterjee, and Satyajit Chakrabarti. “Effect of Public Sentiment on Stock Market Movement Prediction during the COVID-19 Outbreak.” Social Network Analysis and Mining 12, no. 1 (2022): 92. https://doi.org/10.1007/s13278-022-00919-3.
Olson, Emily. “Google Shares Drop $100 Billion after Its New AI Chatbot Makes a Mistake.” NPR, February 9th, 2023. dropped.
xAI. “Announcing Grok.” xAI, November 4th, 2023. https://x.ai/.
Perez, Sarah. “X’s AI Chatbot Grok Now ‘Rolled out to All’ US Premium+ Subscribers, English Language Users Are Next.” TechCrunch, December 8th, 2023. https://techcrunch.com/2023/12/08/xs-ai-chatbot-grok-now-rolled-out-to-all-u-s-premium-subscribers-english-language-users-are-next/.
Thiel, Peter., Masters, Blake. Zero to One: Notes on Startups, Or How to Build the Future. Germany: Crown, 2014.
Heaven, Will Douglas. “OpenAI Teases an Amazing New Generative Video Model Called Sora.” MIT Technology Review, February 15th, 2024. https://www.technologyreview.com/2024/02/15/1088401/openai-amazing-new-generative-ai-video-model-sora/.
Welcome to SOPHIA. My name is Adam Blumenfeld, CEO of CSX Labs— an exploratory research institution dedicated to advancing technology for humanity by creating and contributing to unsaturated (or emerging) technologies.
This publication furthers our mission by combining investigative journalism with peer-reviewed research to create a holistic picture of technologies — while providing quotes and statistics for individual interpretation. The style of reporting is something I wish I had for decisions at my organization.
Although independent, this volume is free to promote reach. We ensure we can provide accurate and transparent reporting free from sensationalism that taints the trust of an article. We may get advertisers in the future to keep us independent, though we will ensure they are independent of the editorial process.
The Generative AI series is our first volume as it is a sensational industry, meaning there is a desperate need for accurate reporting. If you sign up as a member, you will see new articles as they come.
Best Regards,
Adam Blumenfeld