Generative AI Growth and a Test Case for Usefulness
Fred Steube - February 14, 2023
With the populairity of generative AI buzzing our newsfeeds and filling our tech newsletter inboxes, I wanted to see how well ChatGPT performed in terms of generating relevant, usableable content to write a piece on generative AI. I asked ChatGPT to provide me some statistics related to generative AI. Here are the results of ChatGPT responding to my query for an overview and key stats associated with generative AI:
What is ChatGPT?
ChatGPT is a large language model created by OpenAI that can generate human-like responses to natural language queries. Trained on a massive amount of data, ChatGPT has been designed to understand and interpret language in a way that enables it to provide helpful, informative, and conversational responses to a wide range of topics. As a result, ChatGPT is used in a variety of applications, including customer service, virtual assistants, and chatbots, and is available to the public through various platforms and APIs.
Key Generative AI Stats
As of 2021, there were over 20,000 AI-generated artworks sold at auction, with a combined value of over $40 million.
OpenAI's GPT-3 model has 175 billion parameters, making it one of the largest language models in existence.
A study published in 2020 found that 78% of people were unable to distinguish between AI-generated and human-written news articles.
In 2018, researchers at NVIDIA created a GAN (Generative Adversarial Network) that was capable of producing photorealistic images of fake celebrities that were indistinguishable from real ones.
In a survey of 329 data scientists conducted in 2019, 47% said they were currently using generative models in their work.
Google's DeepDream algorithm, which uses neural networks to create abstract art, has been downloaded over 30 million times.
In 2020, researchers at OpenAI created a language model called GPT-3 that was capable of generating coherent and convincing text in a variety of styles and genres, leading to concerns about the potential misuse of such technology.
In a 2021 survey of 300 business leaders, 64% said they were planning to increase their investment in AI and machine learning, with generative models being one of the key areas of focus.
A study published in 2019 found that generative AI models could be used to generate synthetic data that was nearly as accurate as real data, potentially reducing the need for expensive data collection and labeling.
In a 2020 survey of 2,000 people in the UK, 67% said they were excited about the potential of AI to create new forms of art and media.
Then I asked for additional generative AI statistics, and received the following:
In a study published in 2021, researchers used generative models to create a dataset of over 100,000 synthetic images of faces that were used to train facial recognition systems. The results showed that the synthetic data performed nearly as well as real data in identifying faces.
In a 2020 survey of 1,000 people in the US, 71% said they were concerned about the potential misuse of AI-generated content, particularly in the context of disinformation and fake news.
The use of generative models in the video game industry is becoming increasingly common, with companies like EA, Ubisoft, and Nvidia all investing in the technology to create more realistic and immersive game environments.
In a 2019 study, researchers used a GAN to generate synthetic MRI images that were nearly indistinguishable from real ones. The results showed that the synthetic images could be used to train medical AI systems with high accuracy, potentially reducing the need for large amounts of real patient data.
In a 2021 survey of 3,000 business and technology leaders, 61% said they believed that generative AI would have a significant impact on the creative industry, particularly in areas like advertising and marketing.
The use of generative models in the fashion industry is also becoming more prevalent, with companies like Adidas, H&M, and Zara using the technology to design new clothing styles and create digital models of their products.
In a 2020 study, researchers used generative models to create synthetic satellite imagery that was nearly as accurate as real data. The results showed that the synthetic images could be used to train machine learning models for tasks like disaster response and urban planning.
The development of generative models is also being used in the field of music, with companies like Amper Music and AIVA (Artificial Intelligence Virtual Artist) using the technology to compose original music pieces.
In a 2019 survey of 200 senior executives, 74% said they believed that generative AI would be a key driver of innovation in their industries over the next five years.
The use of generative models is also becoming more widespread in the field of natural language processing, with companies like OpenAI and Google using the technology to develop advanced language models for tasks like chatbots and language translation.
When asked for even more generative AI statistics, ChatGPT shared the following:
In a 2021 survey of 1,500 consumers in the US, 60% said they would be willing to purchase products designed by AI if they were aesthetically pleasing and met their needs.
The use of generative models in the field of architecture and urban design is also gaining popularity, with companies like AI SpaceFactory and Urban Logic using the technology to design sustainable and eco-friendly buildings.
In a 2020 study, researchers used a GAN to generate synthetic CT scans of lungs that were used to train a deep learning model to detect lung cancer. The results showed that the synthetic data was nearly as accurate as real data in identifying cancerous cells.
The use of generative models in the field of education is also becoming more common, with companies like Knewton and Carnegie Learning using the technology to develop personalized learning materials and educational content.
In a 2021 survey of 1,000 consumers in the US, 43% said they would be interested in using AI-generated content to enhance their social media presence, such as by using generative models to create customized images or videos.
The development of generative models is also being used in the field of drug discovery, with companies like Insilico Medicine and Atomwise using the technology to identify new drug candidates and accelerate the drug development process.
In a 2020 study, researchers used a GAN to generate synthetic financial data that was used to train machine learning models for predicting stock prices. The results showed that the synthetic data was nearly as accurate as real data in predicting stock trends.
The use of generative models is also being explored in the field of scientific research, with researchers using the technology to simulate complex physical systems and generate new hypotheses.
In a 2021 survey of 1,000 business leaders, 75% said they believed that generative AI would be a key driver of innovation in their industries over the next five years.
The development of generative models is also being used in the field of sports, with companies like Second Spectrum and StatsPerform using the technology to analyze game footage and provide real-time insights and predictions.
