A brief summary of the top AI news stories from February.
The Tension Between Microsoft and Google Continues to Boil Over
This month, Microsoft added AI capabilities to its Bing search engine, Microsoft Edge web browser, and Skype. They all run on a new large language model from OpenAI, which is more powerful than ChatGPT and GPT 3.5. However, the renovated model still contains serious factual errors and has a tendency to be incorrect.
The meteoric rise of ChatGPT has become a solid concern for Google’s dominance in the search market. However, last December, Google’s management clearly articulated that they would not be rolling out their AI chatbot in order to reduce the reputational risks that might follow the release of an unpolished product. The argument was that a startup can afford to be bold, while a giant like Google should be more conservative.
Despite these statements, the company declared a “code red” over the buzz around ChatGPT and put the efforts of many employees (and what’s more, even co-founders, who returned to day-to-day work after years of absence) into accelerating the development of AI products. As a result, in early February, Google piloted Bard, a conversational AI service based on Google’s large language model LaMDA, for “trusted testers” ahead of making it more widely available to the public. Also, under pressure from Microsoft’s success, Google has invested $400 million in the AI startup Anthropic, which is developing a rival to ChatGPT.
“Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence and creativity of our large language models. It draws on information from the web to provide fresh, high-quality responses. Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills.”— Sundar Pichai, CEO of Google and Alphabet, said in a blog post.
Bard’s very first responses have already lacked accuracy, as users caught the chatbot at the false statement that the James Webb Space Telescope took the first pictures of exoplanets, which is not true.
According to Reuters, the case has caused Alphabet to lose $100 billion in the market and sparked a backlash from Google employees on the internal forum Memegon. They have criticized the CEO for rushing to announce Bard, as reported by CNBC.
Meanwhile, the competition between Microsoft and Google is becoming more and more intense. Bill Gates pointed out that Google’s profit from search products is likely to decrease because of Microsoft’s efforts to challenge Google’s dominance in blending cutting-edge AI technology into search.
Meta Creates a Team to Build Generative AI
Meta’s CEO Mark Zuckerberg revealed the company’s plans in his Facebook post. He said that Meta is “pulling together a lot of the teams working on generative AI across the company into one group.” The generative AI team will focus on integrating tools into Meta’s existing products. In the long term, the company will develop “AI personas” that can help users in a variety of ways.
As part of the generative AI strategy, Meta will experiment with chatbots on WhatsApp and Messenger, with images on Instagram, and “with video and multi-modal experiences.” According to Axios, the generative AI team will be led by VP AI/ML Ahmad Al-Dahle.
New AI Technology Behind Stabble Diffusion Allows to Change Visual Style of Video and Movie
Runway, an AI startup for content creators, has introduced a new generative AI technology called Gen-1 that can apply the composition and style of an image or text prompt to the target video.
The approach they used in their research is called video-to-video. The technology has tremendous potential in film and video production, as well as virtual and augmented reality, because it can apply the style of any image or text to a source video, or modify objects in a video given a text prompt, or turn mockups or untextured renders into a stylized and animated video.
AI-generated Images Not Eligible for U.S. Copyrights
The U.S. Copyright Office has denied copyright registration for images created by the AI algorithm Midjourney, Reuters reported. The decision came after author Kris Kashtanova applied for copyright registration for the book “Zarya of the Dawn.” While the text of the book was written by Kris Kashtanova, the images were created from text prompts by the Midjourney algorithm. The regulator accepted copyright protection for the text of the book and the way the author arranged the images and stated that the AI-generated images themselves were not eligible for copyright protection because the process of generating the images was not controlled by the user.
“Person who provides text prompts to Midjourney does not “actually form” the generated images and is not the “master mind” behind them,” — says the U.S. Copyright Office letter, which may eventually serve as a precedent for future legal decisions.
Notion Unveils Its AI Writing Assistant
Notion, the productivity and note-taking app, has launched AI capabilities on its platform. AI algorithms can help to summarize notes, extract key points from text, improve writing style or rewrite text to a specific tone, translate content, correct spelling and grammar, explain concepts, and more. The company suggests treating the AI assistant only as a “thought partner” and warns of its limitations. Notion AI’s output may include incorrect information, harmful content, and bias.
AI features are available to anyone with a Notion account. During the promotion, which runs through April, users will get 20 responses for free; after the promotion ends, they will need to add Notion AI to their plan for $10 per month. Users with annual billing subscriptions will get a 20% discount at the price of $8 per month.
Companies Restrict ChatGPT for Employees
As the popular chatbot continues to take social media and the news by storm, many companies are restricting their staff from using ChatGPT at work, following the suit of educational institutions that set off alarm bells after the bot went viral. To date, JP Morgan, Citigroup, Bank of America, Deutsche Bank, Goldman Sachs, and Wells Fargo & Co are among the financial institutions that have clamped down on ChatGPT due to security and privacy concerns. Banks are not the only companies to have blocked access to the chatbot. Previously, Amazon management has warned employees not to enter confidential information into the chatbot, as did telecommunications company Verizon and technology and consulting firm Accenture.