5 Recent AI Notables
Also, a webinar on AI grading and feedback next Friday.

[image created with GPT-4o Image Generator]
Welcome to AutomatED: the newsletter on how to teach better with tech.
In each edition, I share what I have learned — and am learning — about AI and tech in the university classroom. What works, what doesn't, and why.
Last Time: I shared results from an experiment where OpenAI's o1-pro AI model provided remarkably thorough and helpful feedback on an economics research paper, revealing that advanced AI can serve as a valuable thought partner for us and offering promising applications for research development and student feedback. 👈 click here if you missed it
Yes, I am still alive! For those keeping track, I didn’t send a newsletter the last two weeks. (In short, I was on vacation and then I was sick.) But I am back in the groove now…
Today, I share 5 notable AI developments from recent weeks, plus my commentary on their relevance for higher educators.
I also announce two webinars for April.
Table of Contents

❓ 5 Big Recent Developments in AI
1. OpenAI’s New Image Generator
What Happened: OpenAI integrated a much more powerful image generator directly into GPT-4o, making it the default image creator in ChatGPT. Unlike previous image models, this one excels at accurately rendering text in images, precise visualization of diagrams/charts, and multi-turn image refinement through conversation.
Why It’s Big: For educators, this represents a significant advancement in creating educational visuals, infographics, diagrams, and other instructional materials with unprecedented accuracy and control. It’s not perfect, but you can now quickly generate custom illustrations that accurately display mathematical equations, chemical formulas, or process workflows — previously a significant hurdle in digital content creation — without requiring graphic design expertise or expensive software. This capability dramatically reduces the time between conceptualizing a visual aid and implementing it in course materials. And, perhaps most crucially, the improved text rendering means created materials are more accessible to students, as text within images is now clear and legible rather than distorted or illegible as with previous generators. You don’t need to add labels afterwards, in many cases, and instead can direct for them to be added by ChatGPT.
How to Try It: Free, Plus, Pro, and Team tiers can use it now, with Enterprise and Edu coming soon. To use it, select 4o with the model selector in the upper left corner of your chat screen, then click the “…” button right below the chat input box, then click “Create image.” See below…

2. Google Releases More Educator-Focused AI Courses
What Happened: Google continues to put money into AI education initiatives, including two new courses specifically for K12 (click here) and higher education instructors (click here) on using Google AI effectively in educational settings. These complement their existing resources like the Generative AI for Educators course (created with MIT RAISE), the updated Guardian's Guide to AI, and the Experience AI program.
Why It’s Big: These free, specialized resources arrive just as many institutions are scrambling to develop faculty AI competencies without dedicated training budgets. (Trust me, I know, as funding is often the main obstacle for institutions trying to hire me for PD, trainings, consultations, etc.) For teaching centers and faculty development offices, Google's education-specific courses provide ready-made professional development that addresses pedagogical applications rather than just technical skills. Department chairs can now direct curious but overwhelmed faculty to structured learning paths rather than ad-hoc exploration, potentially accelerating AI adoption across disciplines. Sure, it’s all Google-focused, but it’s better for major tech companies and AI developers to invest in high-quality education-specific AI training than not.
How to Try It: It’s all free, so just click the link above that interests you and go from there! The higher ed one starts with free Google AI functionality and goes to the Premium stuff afterwards. And, at worst, the Premium functionality referenced requires a Gemini Advanced account, but you probably have one for free through your institution or can get one for $10-20 a month (mind you, almost everything can be done in Google AI Studio for free or using the free level of NotebookLM; more from me on the Plus version here).
3. Claude Search
Why It’s Big: For academics and researchers, this means Claude can now help gather current literature, identify research gaps, and assist with building stronger grant proposals, literature reviews, etc. Claude is still behind OpenAI Deep Research and Google Deep Research, but this is an important step in the right direction. Given the frontier capabilities of Claude’s top model (Sonnet 3.7) — I prefer it to o1, o3-mini-high, and Gemini 2.0 Flash Thinking for many tasks — it’s good to see Anthropic begin to expand its functionality with tools like this.
How to Try It: The feature is currently available in "feature preview" to all paid Claude users in the United States (with free tier and international access coming soon). Turn it on with the slider at the bottom of your chat window:

4. Gemini 2.5 Pro
What Happened: Google DeepMind released Gemini 2.5 Pro Experimental, which they describe as their "most intelligent AI model" designed for complex problems. It's a reasoning model that’s currently #1 on the LMArena leaderboard (which measures human preferences), and claims state-of-the-art performance across reasoning, math, and coding benchmarks. For instance, it achieved 18.8% on "Humanity's Last Exam" and 63.8% on SWE-Bench Verified (for coding).
Why It’s Big: Unlike OpenAI's o3-mini-high or Claude 3.7 Sonnet with their 200k token context windows, Gemini 2.5's million-token context window combines with its reasoning power to enable us to complete complex and deep tasks. For instance, you could analyze entire semesters of course materials simultaneously — not just individual lectures or assignments — revealing semester-long progression issues that shorter-context models miss. Grant-writing professors can feed Gemini 2.5 the funding opportunity, their preliminary data, multiple papers establishing the research gap, institutional facilities descriptions, and previous reviewer comments all at once, receiving integrated feedback that shorter-context models simply cannot provide when documents must be chunked or summarized. (I covered this use case with Gemini 1.5 Pro already, and all of my prior discussion still applies with this new, smarter version.)
How to Try It: It’s available now for free in Google AI Studio (select it in the model selector on the right side once you are in the prompt view) and to Gemini Advanced subscribers, with a 1 million token context window (expanding to 2 million "soon").
5. Microsoft’s New Agents
What Happened: Microsoft introduced two AI "reasoning agents" for Microsoft 365 Copilot: Researcher and Analyst. Researcher combines OpenAI's Deep Research model with Copilot's orchestration to help with complex, multi-step research tasks by analyzing work data alongside web information. Analyst, built on OpenAI's o3-mini reasoning model, functions like "a skilled data scientist" to transform raw data into insights using Python (with viewable code for verification).
Why It’s Big: It’s big that Microsoft is taking Deep Research, which is already a beast on its own, and combining it with their pre-existing data organization strengths. To pick one domain for examples, these specialized agents could revolutionize administrative work across higher education, dramatically reducing the time senior leaders spend gathering information for institutional decision-making…
Program directors might use Analyst to transform years of enrollment and outcome data into compelling visualizations that strengthen program review documentation or accreditation materials without requiring dedicated data science staff.
Research administrators could expedite grant applications by having Researcher quickly assemble relevant institutional performance metrics and compliance information alongside funding trends.
For deans and provosts, these tools offer new capabilities to conduct just-in-time analyses of budget allocations, faculty productivity, or student success metrics — turning previously month-long projects into afternoon tasks while maintaining the transparency needed for shared governance through the viewable code.
How to Try It: Both agents will roll out to Microsoft 365 Copilot license holders starting in April through a new "Frontier" program for early access to Copilot innovations.
📝 Two Webinars in April
Due to some scheduling obstacles, I couldn’t host my March webinar. But I’ve got you covered with 2 webinars on the books for April!
1. Friday, April 11th: Using AI for Grading and Feedback
Next Friday, from 12 until 1:30pm Eastern, I’m going to cover all the different ways you can save time and improve your grading/feedback workflows with AI.
As part of this webinar, I will provide a range of options that fit your views on the ethics of using AI for these tasks; regardless of what you think about, say, sharing student work with AI, I will have an AI solution for you to help you do your job better and faster.
Interested? If so, just answer “Yes” in the below poll and I will send you an email with more information and the sign-up link. (Note: the cost is $0 for Premium subscribers and $25 for all others.)
All registrants will receive the recording afterwards, even if they don’t attend live.
Want more information about April 11th's webinar on AI grading and feedback? |
2. Friday, April 25th: Why You Should Be Using Zapier (And How)
Then, on Friday, April 25, from 12 until 1:30pm Eastern, I’m going to cover how to use Zapier (and similar apps) to improve your AI-powered workflows and processes.
If you don’t know, Zapier is a no-code/low-code automation tool that enables you to connect any 2 or more apps together, from ChatGPT’s o1 model to Gmail to Microsoft Excel (and 5000+ more), in a workflow that is automated based on triggers, time periods, etc.
In short, it’s an easy way to get data and software, including AI, to talk to each other in a way that saves you time and headaches.
I would argue that Zapier — or another tool like it, such as Make or n8n — is one of the most important tools you learn to use in the AI era, on a par with LLMs like ChatGPT themselves. I save tons of time with it every week, and it’s key to running AutomatED while teaching.
As part of this webinar, I will convince you that you should use it (trust me, this won’t take long), before pivoting to demonstrating 3 simple ways to use it that can change your life as an educator:
Responding to students’ queries and needs in a dynamic and timely way
Meeting management — before, during, and after
Grading and feedback — never do rote work again, and instead focus on what makes you special as an evaluator
These 3 demonstrations will show you general strategies for using Zapier that you can apply to whatever interests you in automating.
Interested? If so, just answer “Yes” in the below poll and I will send you an email with more information and the sign-up link. (Note: the cost is $0 for Premium subscribers and $25 for all others.)
All registrants will receive the recording afterwards, even if they don’t attend live.
Want more information about April 25th's webinar on Zapier? |
What'd you think of today's newsletter? |

![]() Graham | Let's transform learning together. If you would like to consult with me or have me present to your team, discussing options is the first step: Feel free to connect on LinkedIN, too! |