The Emergence of ChatGPT and Billions in Investment: Propelling AI into the Mainstream in 2023

2023 was a significant year for AI, with the emergence of ChatGPT and the influx of billions of dollars in VC funding. ChatGPT quickly gained popularity, reaching 100 million monthly users within two months of its launch. This rapid growth spurred the entry of numerous startups into the market, offering AI tools for synthetic voice and video generation. Throughout the year, AI made significant progress, surpassing initial doubts about its potential to replace Google search.

Major investments in AI were made by venture capitalists, with Microsoft leading the way by investing $10 billion in OpenAI, which is now valued at $80 billion. Other notable investments include Inflection raising $1.3 billion at a $4 billion valuation and Hugging Face reaching the same valuation. Amazon also plans to invest $4 billion in OpenAI competitor, Anthropic, which has its own conversational chatbot, Claude 2.0.

Despite the influx of funding, not all AI founders had smooth fundraising experiences. Stability AI, which developed the text-to-image AI model Stable Diffusion, faced challenges after its CEO, Emad Mostaque, made misleading claims about credentials and partnerships. Additionally, a Stanford study found that Stable Diffusion’s training dataset contained illegal material.

The AI industry saw the rise of several unicorns, including Adept and Character AI. Companies like Typeface, Writer, and Jasper focused on generative AI for enterprises, enabling tasks like automated email writing and document summarization. Meanwhile, Google launched its own conversational AI chatbot, Bard, and AI model, Gemini, towards the end of the year.

AI’s impact extended to various aspects of life. Concerns arose among teachers about using ChatGPT for cheating, leading to its ban in popular school districts. Doctors and hospitals began utilizing generative AI tools for note-taking, grunt work, and patient diagnosis. Politicians employed AI in their campaigns, and both fake news stories and nonconsensual AI-generated porn became issues of concern.

The rise of AI-generated content raised alarms about the creation of toxic content and its potential exploitation. Low-quality AI-generated content flooded the internet, causing disruptions for freelancers who feared losing their jobs to AI software capable of generating content faster and cheaper. AI chatbots were also used in employee recruitment, raising concerns about biases and risks inherent in the technology.

In response to these challenges, tech giants like Microsoft and Google hired red teams to ensure the safety of their AI models. However, many questions remain unanswered, such as identifying biases in datasets and developing meta AI technologies to regulate AI more effectively.

In 2023, leading AI startups faced copyright infringement lawsuits filed by artists, writers, and coders who claimed that their copyrighted content was used without consent or payment. These lawsuits are expected to lead to nuanced rules regarding the fair use of AI by the U.S. Copyright Office in 2024.

Regulation of AI also took center stage, with the European Union’s AI Act and the Biden administration’s executive order requiring disclosure of large AI models with national security risks. Tech companies generally supported these measures, but startups expressed concerns about potential innovation stifling.

There was also an ongoing debate among AI leaders about whether AI technology should be developed openly or by powerful companies behind closed doors. The safety issues associated with open sourcing AI models sparked differing opinions. Internal divisions within OpenAI became public, resulting in the temporary ousting and subsequent reinstatement of CEO Sam Altman.

The economic aspects of AI remain a key question going into 2024, particularly regarding how AI startups will achieve profitability and generate returns for investors. The reliance on GPUs from semiconductor giants adds to the cost and carbon footprint of training AI models. Smaller, cheaper, and more specialized models are predicted to become more prevalent for the majority of AI use cases in the coming year.