I have had exactly three (3) jaw-dropping AI moments over the past 18 months. One of them just happened. I'll list them here along with their implications (and a few honorable mentions). ****** 1. OpenAI's ChatGPT 3.5 I remember where I was, what I was doing, and my first questions to it. I quickly realized this thing was limitless. An hour later, I texted my boss at NYU Stern School of Business and we agreed right then and there we needed a program to teach our MBAs generative AI in a new way. ****** 2. Anthropic’s Claude Opus In March of this year, I tried out the new Claude version, called Opus. I put in an old newsletter of mine, then uploaded a very rough draft of a new one and asked it to create this newsletter in my voice. I watched it populate and said OMG over and over. It wasn’t just accurate. It actually SOUNDED like me. It made up dumb jokes that I wish I had made up. Moreover, it kept the elements of my rough draft that sounded like me, and rewrote the parts that were clearly not yet polished. I called everyone I knew. Now I saw how Anthropic was going to compete - it wasn’t just on bigger context window or safety. It was imitating brand voice. ****** 3. Google’s NotebookLM This was just in the past week. This is one of these things where you have to hear it to believe it. You’ve probably heard how it can create realistic sounding podcasts out of long boring documents, with two people speaking. But it REALLY SOUNDS like two people are talking. I’m not going to try to explain it, because that would be like trying to explain how cool a card trick was. Go to the website, you can try it for free. Or don’t, I don’t care. But unless you experience it, you don’t get it. What this changed for me was a new way of absorbing information. People love podcasts for a reason - they’re easier to digest - and learn from - than long research documents. Problem is people don’t know how to use it yet. They’re just whooping over it. What you need to do is find an actual use case and show people. That’s what I’m gonna do. ****** HONORABLE MENTIONS: These are honorable mentions because they’re awesome but we don’t have access to them yet, so they’re held to a different standard: OpenAI’s Sora. (Seen demos) Google’s Project Astra. (Played with demo at Google I/O) Meta’s new Orion glasses. (The future, y’all.) >>> ALSO: Check this out: My digital course, Generative AI for Professionals, is now available for enterprise. Learn more here and reach out! https://lnkd.in/ejdb9Nm9
Claudio Truzzi
1mo
Same here for #1 and #3. My #2 was Claude's Artifacts : immediately adopted in my workflow
Conor Grennan I'm with you on each of these. If we condensed the jaw drop window to the past 12 months, I would say it's Claud Opus, Notebook LM and OpenAI's Custom GPT launch in November of last year. These are the three most transformational AI releases of the past 12 months.
Conor Grennan finally Business Development Proposal Generation: Input past successful proposals, client requirements, and industry best practices. Ask NotebookLM to help draft tailored proposals for new opportunities. Market Research: Upload market reports, demographic data, and competitor information. Use NotebookLM to analyze market gaps, potential opportunities, and create targeted strategies for your clients.
LIAM DARMODY
1mo
I haven't tried the NotebookLM yet. NEed to do that this week, clearly. Agree with you re: all the others. Claude Opus is scary good at writing in our voice. Personally, it uses too many chili pepper emojis for my liking in casual writing, but it's easy to understand why :)
For 3, my use case was taking the public-facing NIST ARIA pilot evaluation instructions and pasting them in, generating a realistic podcast of a man and woman walking me through it...uncanny.
Steve Harris
1mo
Conor Grennan - what do you think about committing to use NotebookLM given the 'Experimental' badge? (BTW, I agree with you on how good this tool is - I'm just concerned jumping from Keep to NotebookLM given it seems still Experimental and doesn't have a mobile app)
Paul Durban
1mo
The NotebookLM podcast is freaky and cool. I wish there was a way to train it on pronunciations. I tried to train it on the correct pronunciation of a name, and it made it part of the podcast! "Did you know {name} is often pronounced incorrectly?" Then it comically went on to mispronounce the name thereafter.
Clive Pembridge
1mo
Great moments in the recent history of AI Conor Grennan The leaps in video creation, like the recent PIKA update , are also mindblowing and will no doubt change the entertainment industry and brand promotional campaigns going forward. It seems like for AI the sky's the limit.
Time to level up? Reinvent yourself. Get intentional → 25 Years of leadership, sales, and communication expertise. 👉 Guide. Author. Storyteller.
1moI allowed an AI guru talk me out of trying that notebook. Make your own he said... yep, didn't. I'll circle back. I have created some GPT's that I use to help me. I'm not seeing how Gemini can do that... or the notebook? These are for doing certain actions and writings.