I’ve got this archive of content from two podcasts from long ago, before podcasting was cool and mainstream, where I’m talking with smart and interesting people about the future of education, edtech and teaching and learning in the “digital” age.
In revisiting a few episodes, it ridiculous how optimistic I was that by now we would be having much different conversations about teaching and learning because technology would have disintermediated the existing knowledge-information power paradigm. But no, we’re still having the same conversations because naive me didn’t realize that the status quo has too much mass and too few incentives to be easily overcome. It’s all still here: seat time based learning expectations – check, content as king – check, test scores define success – check, devices as textbook replacements – check, digital literacy talked about but not taught – check, classroom expectations for tech not aligned to outside the classroom reality of use – check, digital divide – check, pacing guides and factory based model – check.
The promise and potential of the return to a more human, personalized and relevant model of education that technology tools make possible has yet to materialize as a systemic path forward. To be sure there are pockets of innovation that exist today, places that are disrupting the status quo and providing next level, future forward learning experiences for kids but drop anyone from the 50’s into the majority of classrooms of today and they’d find the format, routines and expectations reassuringly familiar.
The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn. – Alvin Toffler
Enter AI. AI might just be the thing that can overcome the mass of the status quo in a way that even the disruptive nature the Internet and one-to-one devices couldn’t do in Education (that’s education with a big E as in THE SYSTEM). AI’s acceleration curve is straight up and it’s only a matter of time, so I thought I’d get back to my roots and start playing around with AI tools to see what’s possible for AI to take some content (my old podcasts) and do something interesting with them.
I started with a simple prompt in ChatGPT – Summarize this podcast, to which ChatGPT said it did not have access to the Whisper transcription library. Ok. I asked it how I could get access to the Whisper libraries, and low and behold, it gave me a few options with step by step instructions:
To use OpenAI’s Whisper for audio transcription on your own device, follow these steps:
✅ Option 1: Easiest (No Coding)
Use a free website like:
These let you upload your MP3/WAV file and get a transcript using Whisper—no install required.
💻 Option 2: Install Whisper Locally (For More Control)
Step 1: Install Python
If you don’t have it:
- Download from https://www.python.org/downloads/
- During install, check “Add Python to PATH”
Step 2: Open a Terminal (Mac/Linux) or Command Prompt (Windows)
Step 3: Run These Commands:
pip install git+https://github.com/openai/whisper.git
It may also require ffmpeg. You can install it with:
- Windows (use Chocolatey):
choco install ffmpeg - Mac (use Homebrew):
brew install ffmpeg - Linux:
sudo apt install ffmpeg
Step 4: Run Whisper on Your File
whisper path/to/your/audio.mp3 --model base
This will transcribe your file and save the text as .txt.
Would you like help setting this up on your computer, or do you prefer using one of the websites?
What ChatGPT just spit out in 20 seconds is the kind of thing I could get out of Google in 10-15 minutes of searching technical forums online 20 years ago. Back then search was a superpower, now? AI seems superhuman.
I’m off to install Whisper and see if I can summarize a podcast episode and then, who knows what else I can collaborate with AI on… the possibilities seem limited only by my imagination.
Creativity can solve almost any problem. The creative act, the defeat of habit by originality, overcomes everything. – George Lois
From the Archives: RebootED – Pilot Episode
There is no box 
Leave a comment