generative ai tools
Get up to speed on who is making what in Generative AI. OpenAI is the company with the strongest brand awareness due to their launch of ChatGPT. Microsoft has become their largest shareholder (around 49%) with an investment of $13billion, which is seen as significant enough to trigger a review in the UK from the competition regulator.
Ownership of the data that these tools are trained on is set to become a contentious and already litigious issue as these large companies race to develop their products.
Creator: OpenAI - CEO, Sam Altman
Likely the most well known tool as it was first to launch. ChatGPT is developed by OpenAI and is part of the GPT (Generative Pre-trained Transformer) series. It's a language model designed for natural language understanding and generation, making it useful for chatbots and various text-based applications. A new version is due out in early 2024.
Originally launched as Bard, Gemini is a conversational generative artificial intelligence chatbot developed by Google. A new version was launched December 6, 2023 powered by Google's Gemini LLM. The new version is able to recognize video, images and text inputs.
Axios article on Gemini launch
Anthropic PBC is an American artificial intelligence startup company, founded by former members of OpenAI. It is used by platforms such as Slack, Notion and Quora. Google is a large investor.
DALL-E is another creation by OpenAI. It is a variant of the GPT-3 model and is specifically designed for image generation. DALL-E can generate diverse and creative images based on text prompts. As of writing, the copyright for any DALL-E created images sits with OpenAI.
Creator: Midjourney, Inc.
Midjourney is an image generator that grew rapidly due to it's very simple user interface. Rather than have a desktop application, Midjourney uses Discord where users give their prompts to a bot and the results appear right in the Discord chat. Chaotic, but also simple. They have however announced in December 2023 a new desktop application they are trialing with their high users. Midjourney is currently facing lawsuits from artists as their image 'training data' (the equivalent to an LLM for a chatbot) uses any images it finds online regardless of their copyright status.
Creator: Stability AI - made up of researchers in the AI and machine learning community
Stable Diffusion is open-source code used less by end users than by developers who integrate it into their own image creation products. Stable Diffusion is also facing copyright lawsuits.
Adobe Firefly is an AI-powered design assistant created by Adobe. The first commercially released Firefly model was trained on Adobe Stock images, openly licensed content, and public domain content where copyright has expired and is designed to generate images safe for commercial use.