Innovation Profs - 4/15/2025

Your weekly guide to generative AI tools and news

In this week’s newsletter

Upcoming Innovation Profs Events

👾 Vibe Coding Lunch Club
When: 12:30-1:30 p.m. Central on April 23
Where: Virtual on Zoom
Cost: $30
Attendees will learn about vibe coding, a new trend in generative AI where people are able to code projects with little to no coding skills by simply describing what they want to create. Sign up now.

😎 Hot GPT Summer 2
When: Noon Central on May 14
Where: Virtual on Zoom
Cost: FREE
Our free kickoff to summer is back. Join us for this 45-minute Zoom call to catch up on the latest advancements in the world of generative AI. Sign up now.

Latest Gen AI News

OpenAI debuts its GPT-4.1 flagship AI model

Just yesterday OpenAI launched GPT-4.1, a successor to GPT-4o that has a significantly larger context window (which determines how much information it can keep track of in responding to prompts), is cheaper to run, and outperforms GPT-4o in “just about every dimension.” Two lightweight versions of the model, GPT-4.1 Mini and the even faster and cheaper GPT-4.1 Nano, were also released alongside GPT-4.1. These models are currently only available to developers through the OpenAI API.

OpenAI will soon phase out GPT-4 from ChatGPT

With the launch of new models comes the putting out to pasture of the old ones. In this case, the model in question is GPT-4, which will be removed from the ChatGPT website on April 30. Launched way back in March 2023, GPT-4 was the first multimodal model released by OpenAI. As posted in an update to the ChatGPT release notes, “GPT‑4o is our newer, natively multimodal model. In head‑to‑head evaluations it consistently surpasses GPT‑4 in writing, coding, STEM, and more. Recent upgrades have further improved GPT‑4o’s instruction following, problem solving, and conversational flow, making it a natural successor to GPT‑4.” So long, old friend.

Adobe is building AI agents for Photoshop and Premiere Pro

Agentic AI continues to be a significant trend in generative AI, and now Adobe is planning unleash their own agents. Adobe’s “creative agent” for Photoshop can analyze photos and make and implement suggested edits, while a planned agent for Premiere Pro will allow users to direct agents to make rough video cuts. These agents are not currently available but will be introduced at Adobe’s Max event in London on April 24.

Anthropic rolls out a $200-per-month Claude subscription

The mirroring of products and services between OpenAI and Anthropic continues, as Anthropic has released the Max plan for Claude, a $100-per-month plan that grants users five times the rate limits of the $20-per-month plan and a $200-per-month plan that grants users twenty times the rate limits. Max users also get priority access to new models and features. Recall that OpenAI has its own $200-per-month plan, ChatGPT Pro.

Quick Hits

In the Classroom with the Innovation Profs

Follow along as we teach Drake University’s first ever generative AI course.

This week’s topic: Putting the pieces together
Last week: Gen AI Productivity Tools

Lesson: For the next three weeks, our students will be combining their knowledge and their skills with multiple generative AI tools to see what they can create. This week, they will be combining text and image tools. Next week, they can choose the tools to create something related to the Drake Relays. And the third week, they will combine video and audio tools. Stay tuned to see what they come up with.

Your homework: Check out some of the games our students created during our class on Vibe Coding (and sign up for our Vibe Coding Lunch Club on April 23 to learn yourself):

Tool of the week: Bee

We recently learned about the Bee Pioneer AI device from Greg Swan’s must-read Substack. It’s a wearable AI assistant designed to help you manage daily tasks, remember conversations, and gain personalized insights on your life. The device can be worn as a bracelet or clipped to your clothing.

And for only $50, we had to order one. It arrives later this week, so stay tuned for a full review in a Friday newsletter.

AI-generated image of the week

By now you’ve likely seen people tun themselves into action figures using ChatGPT’s new image generation model. So we decided to share some other options. Just attach a photo of yourself and use the prompts below.

Top-left: Transform this person into a Polly Pocket-style miniature toy figure. The figure should be cute, stylized, and plastic-looking, with a large head, small body, and simplified features. Place the figure on a round plastic display stand, similar to classic Polly Pocket toy bases. Include themed accessories around the figure (e.g., tiny laptop, iphone, microphone, camera). The background should be a soft pastel color, like pink or lavender, to match the Polly Pocket aesthetic. Make everything look like a toy set made of glossy plastic.

Top-right: Transform this person into a Cabbage Patch Kid. Keep their facial features recognizable, but stylize them with the classic Cabbage Patch Kid look: round cheeks, button nose, large expressive eyes, and soft yarn-like hair. Use a plush, doll-like texture for the skin, and dress them in colorful, vintage 1980s-style baby clothes. Place them in a playful setting, like a toy room or a garden with giant cabbages.

Bottom-left: Turn the person in the image into a lego person, sitting at a lego desk, working on a lego computer with lego books and plant sitting on the desk

Bottom-right: Create a realistic, modern action figure of NAME HERE, sealed in its original packaging. He is as a JOB TITLE. The action figure should look like a toy, with clearly visible plastic joints, slightly exaggerated proportions like classic action figures and realistic facial features resembling the photo attached. The packaging design should be bold and colorful with clear text and fun descriptions that reflect his profession, and skills. Include two or three small accessories that match his daily work or hobbies (e.g. XXX, XXX and XXX). Make sure the packaging includes his name and playful copyright writing on the box that reflects him. The entire image should look like a product photo of a sealed toy, well-lit with a full figure and accessories visible inside the plastic casing.

What we found

Generative AI video tool Pika recently announced PikaTwists. How Pika describe the tool: Describe the change you’d like to see and watch your wildest ideas come to life, with all the twists and turns you imagine.