Continuing the AI events summary for April. Without further ado, here's our list of significant AI events:
1. MidJurney Added New Powerful Feature
MidJurney added a new feature to their AI bot called 'describe'. It involves uploading an image to MidJurney (still via Discord), and their AI describes what's in the picture. This has many applications, one of which could be 'reverse prompt engineering.'
More about that feature you can read on their manual: MidJurney
2. Amazon Released Generative AI
Amazon has released its own AI toolkit. Amazon Bedrock, as it's called, is part of the AWS ecosystem and includes tools such as 'text generation', 'chatbots', 'image generation', 'text summarization', etc. Amazon Bedrock utilizes technologies like Stable Diffusion, and models from Anthropic, AI21labs, and Amazon Titan.
You can find a more detailed description on the Amazon AWS website: Amazon AWS
3. Stability AI Launched Their First Language Model
Stability AI released StableLM, an open-source language model available in 3 and 7 billion parameters, with more to follow. It builds on prior open-source efforts, using a 1.5 trillion-token dataset, and aims to make AI technology transparent, accessible, and supportive for a wide range of applications.
Official announcement: Stability AI Website
4. Meta Announced Their Intention to Introduce AI Agents for Millions of People.
Meta aims to introduce AI agents to billions of people, according to CEO Mark Zuckerberg. The company is exploring AI applications across its platforms, including chat experiences, visual creation tools, and eventually, the metaverse, where users can create avatars, objects, worlds, and code.
Source: The Verge
5. AutoGPT
Although the first version of AutoGPT was actually released in March, its true surge in popularity really took place in April.
Auto-GPT, an open-source app created by game developer Toran Bruce Richards, uses OpenAI's text-generating models to act autonomously and interact with software and services online. Despite its limitations and potential risks, the app has become popular for automating multi-step tasks, with new apps like AgentGPT and GodMode emerging to simplify its use.
Github repo: Auto-GPT
6. Elon Musk's New AI Company
Elon Musk (founder of OpenAI) has founded a new artificial intelligence company called X.AI, which was incorporated in Nevada last month. Rumors suggest the company aims to compete with Microsoft-backed OpenAI and will develop a generative AI product using thousands of graphic processing units (GPUs). Musk, who co-founded OpenAI in 2015 but left in 2018, has recently criticized the organization and called for a pause on "giant AI experiments."
Source: The Verge
7. AI Wins Sony World Photography Award
German artist Boris Eldagsen refused his prize after winning the creative open category at the Sony World Photography Award 2023, as his entry, Pseudomnesia: The Electrician, was created using AI. Eldagsen used the image to test the competition and spark a discussion about the future of photography, stating that AI images and photography should not compete with each other. Award organizers claimed Eldagsen misled them about the extent of AI involved in the image.
Source: BBC
8. An AI-Generated Drake Song Viral with Millions of Plays
Music corporations have something to worry about
A new Drake song featuring The Weeknd, "Heart on My Sleeve," has exploded on TikTok and Spotify, amassing millions of listens. The track, allegedly generated by AI, was released by an anonymous producer called "Ghostwriter." While it remains uncertain if the song is truly AI-generated or a marketing stunt, the vocals convincingly resemble Drake and The Weeknd, sparking discussion around the potential of AI-generated music.
9. Open Assistant Release
Open Assistant, an open-source clone of ChatGPT, has been released. It can run on an Nvidia card with 24GB VRAM (3090, 4090) or in Google Colab. It is the first to think and respond reasonably well. It may even perform better, as anyone can participate in collecting an open dataset by registering at Open Assistant and completing simple tasks.
Official website: Open Assistant
10. Italy Has Banned ChatGPT
Italy has become the first Western country to ban OpenAI's popular AI chatbot, ChatGPT. The Italian Data Protection Watchdog ordered OpenAI to temporarily stop processing Italian users' data after a suspected breach of Europe's privacy regulations. The move highlights the absence of concrete regulations for AI, with the European Union and China among the few jurisdictions developing tailored rules for the technology.
Source: Reuters