Why OpenAI’s ChatGPT Is Such A Big Deal
Artificial Intelligence (AI) is becoming increasingly integrated into our daily lives, from smartphones to cars.
One of the most critical steps forward in AI is generative AI, which lets machines make their content. OpenAI, a San Francisco-based startup that made ChatGPT, an AI chatbot that works with text, is one of the companies working on generative AI.
OpenAI's ChatGPT is one of the most popular chatbots in the world. It gained over a million users just a week after it opened up for public testing in November 2022. According to OpenAI's CEO, by the end of January, ChatGPT was receiving around 13 million daily visitors.
While chatbots have existed, most have been tailored to specific tasks. ChatGPT is much more sophisticated, using a language model called GPT 3.5, powered by generative AI. This technology looks through huge amounts of data, like Internet data, Wikipedia entries, and old books, to find patterns and create new content. OpenAI has not revealed the specific data used to train ChatGPT, but it is estimated that the cost of training and operating the tool could reach $3 million monthly, or about $100,000 daily.
A tech reporter, Jordan Novet, said that OpenAI trained ChatGPT on Microsoft's public cloud infrastructure, Azure.
Generative AI is a growing field that has venture capitalists excited. Funding for generative AI companies reached $1.37 billion in 2022. Large companies like Microsoft and Google are developing their language models to power their versions of conversational chatbots, and over 450 startups are working on generative AI.
Big names like Microsoft, LinkedIn co-founder Reid Hoffman, and Khosla Ventures have backed OpenAI's impressive project. Elon Musk co-founded the company in 2015. The company's goal is to ensure that artificial intelligence benefits humanity.
Microsoft has invested in OpenAI since 2019, committing $1 billion to the startup. In January 2023, Microsoft announced a third round of investment in OpenAI, showing how important generative AI is in the tech industry. According to reports, Microsoft is considering adding ChatGPT to its Bing search engine to compete with Google.
ChatGPT is capable of generating text and answering a wide range of questions. Users have put the tool through its paces with prompts ranging from the silly to the practical and creative. However, its popularity has been so immense that such high usage has caused issues and difficulty accessing the service. It's common for users to be met with a message saying that ChatGPT is at capacity and can't handle any more queries.
Despite the capacity issues, ChatGPT has garnered a lot of attention due to its ability to communicate with users in a more human-like way than most chatbots and impressive writing skills. Naomi Baron, who teaches linguistics at a university, was impressed by ChatGPT's basic writing quality. She saw that the tool could write well in grammar and style, better than most of the students she had taught.
Rowan Curran, an analyst for Forrester, described the sudden interest in generative AI as akin to an "iPhone moment," where a product is so revolutionary that it becomes a cultural phenomenon. However, as with any new technology, there are concerns about its use. For example, Jonathan Vanian, a technology reporter, believes that tools like ChatGPT could worsen the fake news problem. He explained that although generative AI has much potential, many aspects must be worked out. For instance, it generates a lot of wrong answers. It is confident, but it can get a lot of the history wrong, which makes it unsuitable for use as a legal research tool.
Another issue with the current iteration of AI technology is that it can be biased or toxic if the databases it pulls from are also biased or toxic. Jonathan Vanian explained that people have already found instances of generative AI generating biased answers, such as associating women with housework and men with more scientific endeavors in certain examples. OpenAI has acknowledged this issue but has long plagued the AI industry, and many companies are trying to evaluate its use.
Cleaning up the dataset can come at a real human cost. A recent investigation by TIME found that OpenAI initially outsourced workers in Kenya to sift through thousands of snippets of texts describing horrific situations, including sexual abuse, violence, and hate speech. The worker's job was to mark these examples as inappropriate so that ChatGPT could learn to recognize and filter out harmful content. According to the TIME investigation, a number of these workers were traumatized by their jobs. CNBC asked OpenAI what it thought about what happened, but the company didn't answer.
Even though generative AI like ChatGPT still has a long way to go, predictions about how technology will influence society are already swirling. One prominent discussion centers around generative AI's propensity to replace some human workers. Experts say journalism, law, and translation are among the industries that may be most affected by ChatGPT. Programming may also change, with companies already training generative AI models to write basic code.
Linguistics professor Naomi Baron believes that tools like ChatGPT have the advantage of customizing briefs for a specific judge by taking all of the data from their past opinions and seeing which briefs succeeded in winning the case. But not everyone shares this outlook. Jordan Novet, a journalist who writes about technology, thinks that generative AI is a step forward in how people can think about what AI can do in the world of language.
However, society has already had to adapt. Some school districts around the country have banned ChatGPT because they are afraid that students will abuse it. Rowan Curran explains that while ChatGPT may take away some of the really annoying tasks that were part of our job but that we didn't really want to do, like some of the data cleaning and data prep tasks, it will ultimately create more opportunities for knowledge workers in the long run than the ones it will prevent.
To better understand ChatGPT and how it works, Magdalena Petrova, a CNBC reporter, turned to the chatbot, using text-to-speech software to give It a voice.
Petrova asked ChatGPT about its capabilities, knowledge, and limitations with a series of interview questions answered in a one- to two-sentence interview format.
ChatGPT described itself as a large language model trained by OpenAI capable of generating text, answering questions on various topics, and producing creative writing and text summaries. Its knowledge is based on a massive corpus of text that includes books, articles, and websites. This training enables ChatGPT to understand and generate text on many different subjects.
ChatGPT said that its answers are limited by the data it was trained on, and it may not always give complete or correct answers. Additionally, as an AI model, it may make mistakes based on its predictions. Remembering these limitations when using ChatGPT for research or informational purposes is critical.
Another area of concern is the potential for misuse. When asked about the possibility of providing instructions for illegal activities such as bomb-making, ChatGPT clarified that it cannot provide such information. It's important to remember that ChatGPT is a powerful way to find information, but it must also be used responsibly and ethically.
When asked about AI's effect on the world, ChatGPT said it could have a big effect on many industries and fields, but it was important to use it responsibly to make sure its effects are good.
ChatGPT is a sophisticated chatbot that could change how we find and use information. But, as with any technology, it's important to consider its limits and how it could be used wrongly. Even though OpenAI turned down an interview, Petrova's direct interview with ChatGPT gave important information about its strengths and weaknesses and a look into the future of AI and how it might affect different industries.
Generative AI is an exciting field with endless possibilities, and ChatGPT is just the beginning. As technology continues to evolve, so will the capabilities of generative AI, but it is up to individuals to use it safely and responsibly.