In the beginning of this year, we shortly informed you in our newsletter about the ChatGPT and Open AI. Since than, we have been monitoring the situation and took notice of many new articles published in, among others, Financial Times. Let us give you an overview of ChatGPT and what it means for the investment world.
Generative AI refers to artificial intelligence systems that can generate new content or data. These systems are trained on large datasets to learn patterns and characteristics of the data. Then they can then use this knowledge to generate new, original content that is somehow similar to the training data.
OpenAI is a research organization that focuses on developing artificial intelligence technologies and applications. The aim of this limited-profit company is advancing AI in a responsible and safe manner, GPT is only one of notable projects of the company, among others is DALL-E, a generator of images based on text prompt.
GPT (short for “Generative Pre-trained Transformer”) is a deep learning model that has been trained on a massive text dataset and so is able to generate human-like text when given a prompt. It can be used for a variety of language tasks, including language translation, article summarization, and new text generation.
Do you want to know how well the GPT works? The previous three paragraphs were written with the help of generative AI! OpenAI made ChatGPT available to the public in order to collect more data to train its language and clear it from bugs or imperfections.
According to data from PitchBook, venture capital investment in generative AI has increased by 425% since 2020, up to 2.1bn USD last year. According to Financial Times, investors say that this technology could be a similar transformation as mobile was 20 years ago.
At the end of January, Financial Times informed about a multibillion-dollar investment of Microsoft in OpenAI. Details around this transaction as well as precise financial amount have not been disclosed, we only know that it was an investment of a few billions of dollars in a “multiyear” agreement. Reporters in the article informed that OpenAI was looking for $10bn from Microsoft at $29bn valuation. “The huge boost of capital for OpenAI, the San Francisco-based developer behind chatbot ChatGPT and image generator Dall-E, is symbolic of investor enthusiasm for “generative AI”, which is capable of generating high-quality output based on just a few words or sentences of human instruction.” quoting from the article. Microsoft started investing into OpenAI in 2019, when they invested its first $1bn as a beginning of the partnership to help with the commercialisation of the AI technology. “Microsoft shares our values and we are excited to continue our independent research and work towards creating advanced AI that benefits everyone,” said Sam Altman, the chief executive of OpenAI.
Excitement around ChatGPT started in November overnight, when the company gained over 1mn users in the first week after its launch. Richard Waters and Tabby Kinder wrote in their article in the Financial Times from the beginning of January that venture capital investors are seeking investments in the latest AI trends.
In the beginning of February, George Steer wrote about artificial intelligence stocks soar on ChatGPT and other generative AI models’ hype. Analysts now warn of a bubble in the market. After chief executive of BuzzFeed, Jonah Peretti, said in late January that “AI inspired content” would be a part of the company’s website later this year, BuzzFeed’s stock rose 150% in a single day. Analysts at Morgan Stanley said that generative AI technology was showing “all the usual hallmarks of hype: social media echo chambers, exponential venture funding and polarised media … During our travels around the US last week, ostensibly to discuss decarbonisation and reshoring, conversation regularly turned to ‘and what do you make of ChatGPT?’”
The hype around ChatGPT is undeniable. According to UBS is ChatGPT is gaining around 100mn monthly active users in January, just two months after the launch. “In 20 years following the internet space, we cannot recall a faster ramp in a consumer internet app,” the UBS said. “It took TikTok nine months after its global launch to add 100mn incremental users and it took Instagram 2.5 years.”
Arjun Neil Alim informed in his article about using ChatGPT to help the publisher of the Daily Mirror and Daily Express newspapers to write local news. The company has set up a working group to explore the potential and the limitations of the tool and its use to help human reporters covering routine stories such as local traffic or weather. Francesco Marconi, co-founder of computational journalism company AppliedXL said for Financial Times that ChatGPT can have a supporting function for reporters, can help journalists edit text but can’t generate original reporting.
However, there is also a negative perspective in that this innovation can help spread misinformation. The technology still needs human intervention, as Madhumita Murgia wrote in his article about US-based sci-fi magazine Clarkesworld. Since December, when ChatGPT was released to the public, the number of stories written by AI has been increasing. In February it was more than 500 and Neil Clarke, the founder and editor, closed the submissions, since it was impossible to filter and deal with the volume of content.
Moreover, the coding Q&A website Stack Overflow was forced to stop using ChatGPT for generating responses, when it was shown that ChatGPT fills up their forum with misinformation. In February OpenAI released an experimental tool which should detect AI-generated content, after educators were concerned about plagiarism at schools and universities. However, the success rate of this tool is just 26%.
In the end, as Darren Dodd mentioned in his article in the Financial Times, “technological developments are running ahead of ethical considerations”. Many questions are around copyright as well as about using this technology to cheat at schools. There is already a reported case where some students on an MBA course at Wharton, one of the most prestigious and oldest US business schools, were found to use ChatGPT. Concerns came not just from academic world but even from Vatican, “where tech correspondent Madhumita Murgia witnessed a remarkable meeting of tech and religious leaders who agreed on a set of principles, known as the Rome Call. They include making AI systems explainable, inclusive, unbiased, reproducible and requiring a human to always take responsibility for an AI-facilitated decision”, as informed Financial Times last week. We will monitor this market in detail to inform you further.