[Music] Our newest flagship model, GPT-40 Gemini, is delivering our most intelligent AI experience. With the launch of Amazon Nova, we are introducing more than 50 new products. We're going to try something that, as far as we know, no tech company has done before.
Google has been at the forefront of AI better than any other; it is incredibly advanced and more capable than any other model. Gemini is our largest and most capable. 2.
0 enables a new frontier in intelligence. As a product marketing manager at Google, believe me, I feel that pressure to keep up with AI more than most. Initially, I struggled because, one, there's just so much noise out there, and two, I felt held back by my non-technical background.
After two-plus years of consuming newsletters, podcasts, courses, embedding AI tools into my workflow, and even running workshops to test my own knowledge, I can confidently say the problem is not that you're behind; the problem is that nobody has shown you a clear and actionable roadmap to navigate all of this. So in this video, I'll break down the three biggest challenges holding people back in AI and the exact strategies I use to overcome them. Let's get started.
Challenge number one: AI tools paralysis. Product launches, new feature announcements, and benchmark-pissing contests make up most of the noise around AI, and all this hype makes it nearly impossible to identify which tools can actually boost our daily productivity. Because let's be real: new features and impressive benchmark scores mean nothing if they don't translate to real-world impact.
Here's a perfect example: when OpenAI released their GPT-3 model, it dominated every benchmark test against their previous version, and users reported outstanding results. So naturally, everyone should immediately switch to the newer model, right? Well, no.
The older model was still 30 times faster and cost just a third of the price. No one would wait an entire minute for ChatGPT to edit a single email. This confusion only gets worse when you look at the bigger picture; every major tech company claims to have the best AI tool, and each platform offers multiple models.
And that's just for text generation! We haven't even touched on text-to-image, text-to-video, and text-to-audio tools. This is how I initially fell into the tools paralysis trap.
I sort of lost the motivation to learn and use AI tools because I thought, "Well, if a new tool is just going to make this one obsolete, why bother? " Luckily for me, I came across this action plan checklist when I was taking one of Google's AI courses. This checklist inspired me to develop what I call the Minimum Viable Toolkit strategy.
Here's the core idea: even if every AI tool out there is incredible—which they're not—we can't possibly become proficient in all of them. Therefore, we should ask ourselves, "What's the smallest set of AI tools that will cover all my essential needs? " I'm going to oversimplify a bit here, but in practice, this is a process I follow: Step one: Spot your recurring need at work.
I needed a faster way to research digital advertising trends. ChatGPT hallucinated too much, and Gemini's research mode was too slow. Step two: Find your tool.
After testing both Consensus and Perplexity, Perplexity won me over with its speed and user-friendliness. Step three: Master one tool. I committed to using Perplexity for all my research until it became second nature.
After consistently meeting my needs, it earned a permanent spot in my toolkit, and I even published an entire tutorial video on Perplexity, which I'll link below. Here's a contrasting example: like many of you, I wanted a quick and reliable way to create charts for work presentations, and I discovered Napkin AI, which promised to turn text into stunning visuals. While I love the concept and I desperately want them to succeed—especially since the founders are ex-Googlers—the current iteration can't produce the sophisticated graphics I need, so it doesn't make the cut for now.
By the way, if you want to skip the trial and error, you can just grab my free AI toolkit, which is designed to cut through the noise and help you master essential tools and workflows. I'll leave a link down below. To be clear, I am not saying you should ignore every new product announcement.
I am saying that most new tools are either too specialized or too unpolished for everyday use. The best strategy is to stay focused, identify your key needs, and slowly build out your Minimum Viable Toolkit. Challenge number two: Death by Prompts.
Once you identify a few AI applications that fit your needs, there's still this invisible barrier that prevents you from using them consistently, and I call this problem "death by prompts. " Here's what I mean: I have this incredible prompt that transforms my writing into clear and concise copy. I use it for everything: emails, reports, YouTube scripts, pickup lines for Bumble… I'm obviously kidding about that last one—I don’t get matches.
But there's absolutely no way I'd be using this 20-plus times a day if I had to type it out every single time because the friction of typing this 20 times a day is just too much. The good news is there are two surprisingly simple solutions to this problem. The first is to install and start using a text expander.
Check this out: my talking points for this business planning slide are too wordy. I copy the text, type “::chpt concise clear,” and boom! The entire prompt appears instantly.
I make a few edits, paste in my talking points, and the output is much better than what I had. Originally, I personally use a paid app called Alfred, but free alternatives include Raycast for the Mac and BeefTex for Windows. Use it every day for a week and see just how much time you save.
Moving on to strategy number two: embedding prompts directly into your workflow. The key insight here is to organize information based on where you will use it, not where you found it. For example, when we schedule a health checkup, the smart move is to add the “no breakfast” and “bring ID” reminders directly in the calendar event, rather than our Notes app.
The same principle applies to AI prompts. In my case, I have a weekly calendar block where I need to summarize project updates for leadership. Instead of trying to remember the prompt to use each time, I've included a hyperlink to a specific prompt within the calendar event description, along with my text expansion shortcut, all designed to decrease the friction as much as possible.
At this point, you might be thinking, “Wait, Jeff, where does that link point to? ” The answer is: all my prompts are stored in a prompts database within Notion. This is out of scope for this video—subscribe for a full breakdown—but what matters is having one central location for your prompts and then linking to them from your actual workflows.
Another real-world example: after every marketing campaign, we receive feedback from hundreds of clients, but the formatting is usually all over the place. So, within my project management spreadsheet for that campaign, I always include a link to a prompt that transforms raw feedback into a clean and consistent format. As you can hopefully see by now, using text expanders and embedding prompts into your workflow are essential to removing all the little friction points that usually get in the way.
Challenge number three: update suffocation. So far, we've covered strategies that help us identify useful tools and decrease the friction of using those tools. But every day, we still face what I call update suffocation—an overwhelming flood of information that just never stops, and this causes a major roadblock in our learning journey: decision fatigue.
A colleague at Google recently said to me, “Jeff, I'm not even 100% comfortable with Gemini yet, and now they’re announcing all these AI agent capabilities. Should I start learning those first or stick with prompting? ” The solution I found is a simple two-step system I call the impact loop.
Here’s how it works: Step One is Learn. Unfortunately, to stay competitive, we do need to consume AI updates daily, but the key insight here is to be ruthlessly selective and just follow one to two trusted curators who consistently deliver the most important AI updates. If you prefer to receive updates over email, I recommend subscribing to these six popular newsletters, then unsubscribing from the ones you don’t vibe with after a few days.
This way, you're left with one to two subscriptions. If you prefer listening to updates, I recommend the AI Daily Brief podcast, but I personally prefer email updates since it's just easier to take notes. Regardless of the consumption medium, the idea is to build a habit of spending 5 to 10 minutes a day staying current because this naturally leads to Step Two of the impact loop: taking action.
Block out a specific time every week. For me, it's 1 hour every Saturday to actually experiment with one thing you learned—not everything, just one thing. For instance, when I read about ChatGPT's Notion integration on a Friday, I added a task reminding myself to explore this new feature.
Come Saturday, I spent 30 minutes testing the integration and found it super helpful, teaching me how to write Notion formulas. Another example would be when I learned that Google released an extensive prompt library. I bookmarked the web page, added a task for myself, and on Saturday, I explored and tested prompts related to my work.
As you can probably tell, the key to the impact loop strategy is consistency. Even if you can only spare 30 minutes a week, taking action and implementing something you've learned is infinitely more valuable than hours of passive consumption. So, are you falling behind with AI?
No. The real challenge isn't about catching up; it's about building sustainable systems. By following the three strategies in this video, I promise your relationship with AI will change.
Instead of chasing every shiny new tool, you'll build a focused toolkit that handles 80% of your needs. Instead of struggling with prompts, you have them seamlessly embedded into your daily workflow. And instead of drowning in updates, you'll be confidently learning and implementing AI at your own pace.
If you found this helpful, check out my AI playlist next. See you on the next video! In the meantime, have a great one!
