Harvard University is using an artificial intelligence tool called ChatGPT to help teach beginner computer science courses. The tool, known as the CS50 Bot, provides personalized assistance to students with coding challenges, giving them immediate feedback and support. This AI tool aims to enhance the learning experience and free up teaching assistants’ time.

The article discusses the race to build an operating system for generative AI and its impact on enterprises. It highlights the challenges faced by enterprises in adopting generative AI and the need for a new infrastructure to support the intelligence and autonomy of generative AI applications. The article also explores the four main layers that enterprises need to consider when building an operating system for generative AI. Additionally, it mentions the emergence of open software frameworks and platforms that enable developers to create more intelligent and autonomous generative AI applications. The article concludes by emphasizing the strategic advantage that enterprises can gain by mastering generative AI and the opportunity to learn more about it at the VB Transform event.

This article explores the features and enhancements of Midjourney version 5.2, an AI art generator. It highlights the Zoom Out function, which allows users to expand an image’s canvas without affecting the original content. The article also discusses other notable features of Midjourney 5.2, such as Variations Mode, Stylize command, Weird command, Shorten command, Turbo mode, and Details View. It provides insights on how to maintain character consistency across different scenes and offers tips on maximizing the use of the Weird command. Additionally, the article mentions the Shorten command as a prompt optimizer and Turbo mode for faster image rendering. The Details View feature is explained as a tool for understanding the AI’s keyword ranking.

This article discusses Auto Gmail, a generative AI assistant that can automatically respond to emails and create drafts for inspection. It explains how Auto Gmail was created to address the problem of handling email messages in a timely manner, especially in the face of spam and promotional emails.

AI is so powerful that it will inevitably destroy the world—at least, that’s what the people who sell AI software keep saying, and I can’t think of any reason why they might lie about how amazing they are. Still, I can’t help but wonder: What is AI useful for right now, before it ends civilization? I’ve done some experimenting and talked to my friends on LinkedIn and Mastodon. Here’s the best use cases I could personally find.

The South Australian government has announced a trial of AI technologies in eight public high schools. This trial, known as the SA school trial, aims to explore the use of AI to support student learning and understand the benefits and risks associated with these technologies. The trial will involve the use of an AI chatbot called “Edchat,” which has been developed by the state’s education department and Microsoft. The government will assess the results of the trial to determine whether to adopt the tool for other schools.

This article discusses the four types of prompts that entrepreneurs can use with large language models (LLMs) like ChatGPT to maximize their effectiveness. The author, Lasse Linnes, provides examples and explanations for each type of prompt: information prompts, creative prompts, instructional prompts, and problem-solving prompts. By understanding these different types of prompts and how to use them, entrepreneurs can leverage AI tools like ChatGPT to enhance their business operations.

Generative AI is utterly reliant on the sustenance it gets from the web: Computers mime intelligence by processing almost unfathomable amounts of data and deriving patterns from them. ChatGPT can write a passable high-school essay because it has read libraries’ worth of digitized books and articles, while DALL-E 2 can produce Picasso-esque images because it has analyzed the entire trajectory of art history. The more they train on, the smarter they appear.

After six months, OpenAI has yet to train its next-generation language model, GPT-5. While the industry waits for its release, improvements to ChatGPT are expected with the release of GPT-4. GPT-4 has notable upgrades, including the ability to “see” through images and graphs, and broader general knowledge. GPT-5 is expected to have multimodal capabilities, be more truthful, and bring us closer to artificial general intelligence. GPT-4 is expensive to use, but its cost may decrease once GPT-5 is released.