Innovation Profs - 4/2/2024

Your weekly guide to generative AI tools and news

Sign up for AI Lunch Club

Want to go deeper into the subjects we talk about in our workshops? We’re doing just that in our new AI Lunch Club virtual events. These will take place over Zoom from 12:30-1:30 p.m. Central:

  • April 3: Go AI Yourself: Create an avatar and make it talk.

  • April 8: Advanced Image Prompting: Take a deep dive into image creation.

  • April 10: Next Steps in Prompting: Advanced prompting techniques for LLMs.

Sign up for events here. Use the code SAVE10 to save $10 off the price for each event. Or bring 3+ people and save more.

Generative AI News

Microsoft and OpenAI reportedly building $100 billion secret supercomputer to train advanced AI

Microsoft and OpenAI are reportedly partnering on a 100 BILLION DOLLAR supercomputer—codenamed Stargate—for training AI systems, with an anticipated launch by 2028. Microsoft neither confirmed nor denied the report, but responded with the following statement: "Microsoft has demonstrated its ability to build pioneering AI infrastructure used to train and deploy the world’s leading AI models. We are always planning for the next generation of infrastructure innovations needed to continue pushing the frontier of AI capability."

“The king is dead”—Claude 3 surpasses GPT-4 on Chatbot Arena for the first time

Chatbot Arena, a leaderboard for AI researchers to rank the abilities of various large language models, has a new top model: Claude Opus has unseated GPT-4 from the top spot, which has held that position since May 2023 (just one week after the Chatbot Arena was rolled out). The smaller Claude models, Claude Sonnet and Claude Haiku, have also recently risen up the leaderboard.

X's Grok chatbot will soon get an upgraded model, Grok-1.5

New and improved versions of various large language models continue to be rolled out. Next up in the queue: X’s Grok-1.5 is set to be released in “the coming days.” Apparently Grok-1.5 will be significantly better at reasoning tasks compared to Grok-1 (which was released as open source just a few weeks back).

Amazon invests another $2.75B in Anthropic — reportedly ‘largest’ in company history

In addition to its recent $1.25 billion investment in Anthropic, Amazon added $2.75 billion more, which is reported to be Amazon’s largest venture investment. As we noted in an earlier newsletter, Amazon has been hosting versions of Claude 3 on its Bedrock platform, which allows users to work with a range of different large language models. As announced by Amazon, “Anthropic’s visionary work with generative AI, most recently the introduction of its state-of-the art Claude 3 family of models, combined with Amazon’s best-in-class infrastructure like AWS Tranium and managed services like Amazon Bedrock further unlocks exciting opportunities for customers to quickly, securely, and responsibly innovate with generative AI. Generative AI is poised to be the most transformational technology of our time, and we believe our strategic collaboration with Anthropic will further improve our customers’ experiences, and look forward to what’s next.”

Databricks launches DBRX, challenging Big Tech in the open source AI race

On the open source front, global software company Databricks launched a new open source large language model DBRX, which outperforms other leading open source models. Although DBRX cannot stack up to the performance of GPT-4, its performance does compare favorably to that of GPT-3.5, but it was trained at a fraction of GPT-3.5’s cost. In a press release, Databricks CEO of Ali Ghodsi stated, “While foundation models like GPT-4 are great general-purpose tools, Databricks’ business is building custom models for each client that deeply understand their proprietary data. DBRX shows we can deliver on that.”

Quick Hits

Tool of the week: Elevenlabs

We talk about Elevenlabs in our presentations, but we haven’t mentioned it much in this space. It’s an AI audio tool with an impressive list of things it can do. Use Elevenlabs to:

  • Generate realistic voices in different languages

  • Clone a voice with as little as one minute of audio

  • Generate sound effects

  • Translate audio and video to 29 different languages

You can get started for free, but will need a paid account for some features such as voice cloning.

AI-generated image of the week

We’ve been testing out custom GPTs and creating avatars in advance of tomorrow’s AI Yourself lunch and learn (still time to sign up here). This avatar of Professor Snider was made with a GPT called PixArt Avatar.

Prompt: No prompt used, but the GPT has built-in prompts to get a Pixar character style.

Generative AI tip of the week: Getting your company on board

Generative AI is a transformative tool, but not everyone in your company might be ready for it. Some tips to bring people along:

  • Approach it as a tool to do your jobs more effectively, not to replace your jobs

  • Create a culture of curiosity and collaboration to be more accepting of trying new tools

  • Find champions in the company who will try the new tools and help others get up to speed when the time is right

  • Pay for the latest versions of tools so you’re not using outdated technology

Get starting with Generative AI

New to generative AI? Here are some places to start…

What we found

Deepfakes are all around us. Here’s a roundup of some recent ones found on X to help you learn to spot the fakes.