While the AI Avatar Helen Todd was certainly engaging, the real Helen was a great speaker, filled with buoyant optimism about all the amazing ways to use this technology. She’s definitely one to watch, and I recommend checking out her “Creativity Squared” weekly podcast.
2. The introduction of reasoning to ChatGPT with the new ChatGPT o1 update
This wasn’t a planned session. It was more like a seismic quake rolling out right before Paul Roetzer, the founder of the Marketing AI Institute, was taking the stage to begin his closing remarks on this final day of the conference. Attendees in the main ballroom twittered with anticipation as Paul tried to find his footing, admitting to the waiting audience that he needed to throw away any idea of prepared remarks in the face of one of the most dramatic releases in the history of artificial intelligence. But with a characteristic laugh at the absurdity and backed by his incredible knowledge as an “AI insider,” he dove in and explained what had happened: ChatGPT had just learned to reason. No longer limited by just scraping the internet, organizing the existing knowledge it has found and returning a nicely organized answer, it can now reason its way to a new conclusion that it has derived itself. With this functionality, it can have a conversation with us to work through complex problems and find our way to new ideas and new solutions.
This is a huge leap forward and will most certainly cause an enormous acceleration of AI technology that can’t be underestimated.
Of course, humans will still need to ensure the logic and resulting concepts are valid and meet the common sense rule, but there should be fewer hallucinations as the ChatGPT o1 model (you’ll see it introduced as “preview” and “mini” at first) since it employs longer strings of “thought” as it formulates a response. According to OpenAI’s release notes on September 12, 2024, “Learning to Reason with LLMs,” it’s expected to particularly excel in topics related to math (equations), coding, data analysis, and science. ChatGPT 4o still has an edge in personal writing though, with access to the live internet and its ability to handle images.
There is still a lot to learn about OpenAI o1, especially since the initial preview and mini releases are the least powerful version of the technology. The actual “model o1” will be dramatically more powerful, and I would assume have access to the live internet at that point. For Paul’s latest thoughts on this release, I encourage you to listen to the Marketing AI Institute’s “The AI Show” podcast episode 115 (Sept 17, 2024).
3. The AI Action Plan
In closing, we were given insight into an “AI Action Plan,” which is something any company that wants to really benefit from AI needs to put together. Given the importance of this technology, I can’t imagine a reason NOT to put a plan in place to adopt AI since the alternative would be equivalent to deciding not to sign up for the internet back in the AOL dial up days.
Everyone’s plan will be different, but my take on a couple “must have” essentials includes:
- Create a learning plan for your team to give them exposure to the tools, and require everyone to get some hands on experience. This isn’t something for the “Back Burner” (a dreaded phrase here at Front Burner Marketing, as I’m sure you can imagine!). Do it today. This will need to be a strategy for continual learning, too. If you’re going to keep up with the pace of knowledge going stale in months vs. years, you’ll need to have a plan in place (sign up for MAICON 25 as soon as it opens!).
- Review your business to determine where the different types of technology fit, particularly within your marketing tech stack. Remember to look at your existing applications as well, since most of them are rolling out new features that are AI enabled. What problems could this help you solve? Where can this save you time or eliminate drudge work?
If all of this is a bit overwhelming and you can feel yourself giving in to the inertia that slips projects like this onto the back burner, I encourage you to take a look at our “AI Ignitor” program here at Front Burner. We created it just for this purpose, to provide an inexpensive way for our clients to get our assistance in reviewing their marketing tech stack to look for opportunities to on-ramp into the AI world. Sometimes it just helps for someone to take a fresh look at things and help you figure out where to get started, and we have several easy deliverables that you can choose to receive as a result of the assessment. So, just sayin’, it’s worth the small expense to gain a huge knowledge advantage and get this moved to the Front Burner!
Key takeaways:
- Virtual avatars are expected to be a $270 billion market by 2030, offering a wide range of uses like SEO videos, customer service, social posts, and online training.
- Creating a personalized digital avatar can save time and reduce production costs while maintaining authenticity with ethical guidelines like consent and transparency.
- The ChatGPT o1 update introduces reasoning abilities, allowing AI to solve complex problems and offer original conclusions, a major advancement in AI technology.
- ChatGPT o1 excels in tasks related to math, coding, data analysis, and science, and will continue to evolve with more powerful releases.
- Businesses should develop an AI Action Plan to integrate AI into their operations, ensuring teams have hands-on experience and continually update their knowledge.
- AI-enabled tools can help businesses save time, reduce repetitive tasks, and improve efficiency by incorporating AI within existing marketing tech stacks.
COMMENTS