Generative AI made its public debut in 2022—it could be an internet earthquake in 2023

The trend: In 2022, Big Tech unleashed the disruptive generative AI technology into the public domain.

DeviantArt saw a 1,000% increase in AI-generated images uploaded to its platform between August and November, per Gizmodo.

  • Building on the success of OpenAI text-to-image generator DALL-E, released in 2021, Big Tech companies have embraced generative AI as a potential money-maker.
  • OpenAI's other big release—ChatGPT—is one of many powerful AI systems unveiled this year that are raising questions about the future of work, companies, and the ethics of commercial AI.
  • Microsoft began integrating generative AI technology powered by DALL-E 2 into its Bing search engine, Edge browser, and a new Microsoft Designer app for Office.
  • Google announced its text-to-video program, Imagen Video, which the company is pairing with its other text-to-video system Phenaki to create longer videos from longer written prompts.
  • Google also released an AI art tool for consumers through its Test Kitchen app.
  • Building off its earlier Make-A-Scene generator, Meta’s AI division unveiled its Make-A-Video tool that can create short video clips from text prompts.
  • NVIDIA also entered the generative AI arena with its eDiff-I, which creates larger, high-resolution images with less processing power.

Much more than hype: Unlike the metaverse that failed to materialize this year, decades of AI research led to the debut of a technology that doesn’t merely analyze data—it creates content like images, videos, blog posts, code, podcasts, and music that didn’t exist before.

  • Its potential to augment digital content creators’ and advertisers’ work led Sequoia Capital to predict that generative AI will “generate trillions of dollars of economic value.”
  • Companies like Shutterstock are rushing to be at “the forefront of AI art” by integrating DALL-E 2 into its content platform, planning to make it available to stock image customers globally.
  • Such moves have raised concerns about the ethics of commercializing AI art that’s based on models trained by human-made art.
  • For this reason, when DeviantArt released its DreamUp AI art generator based on the Stable Diffusion AI model, it set it up to tag AI images and credit human authors involved in its production. Artists also have the option to bar the AI from using their work.
  • Yet so far, giving human creators credit seems to be the exception, not the rule.
  • In November, a programmer filed a lawsuit against Microsoft, OpenAI and GitHub, alleging that the companies trained the generative AI system Copilot on licensed open-source code scraped from the internet with no credit given to the authors.

2023 predictions: Expect a tsunami of AI-generated content to flood the internet in 2023, as companies, including those within the gaming industry, get bolder about deploying commercial AI art tools.

  • Although the content may be helpful to those who want to quickly generate content without paying human designers, it will also likely amount to a new source of spam, misinformation, and fraud that could overwhelm platforms like Google Search.
  • Meanwhile the outcome of the case involving Copilot, and possible others, could set a precedent for how AI content is regulated by copyright laws.
  • In addition to possible legal consequences, expect social pushback against generative AI systems that create controversial or false content like Meta’s Galactica AI.
  • Despite challenges, generative AI might soon become a financial lifeline for the tech industry as it suffers through its recession.

"Behind the Numbers" Podcast