Jerry Foster

K-12 education: Expected to defend against nation-state attacks... with the same budget we use to buy glue sticks and crayons. Cybersecurity is no longer optional — it's critical infrastructure. Yet public schools are funded like afterthoughts, even as the threats get bigger and scarier. Vendors: your tools are amazing. Your prices reflect your worth. But without investment in K-12 security, even the best solutions will sit out of reach for too many of us. We don't need charity. We need recognition. We need funding. We need allies who understand that when a school network falls, it’s not just data — it’s lives. So until then, we’ll keep looking for grants... and maybe a few miracles. #SecureK12 #PublicEducationFirst #FundingMatters #CyberSecurity

160

11 Comments

My experiments bring all the devs to the yard, And they’re like, “oh wow, this is worthless!” Sometimes experiments are successful, and other times times you just embarrass yourself. The trouble is that we're often ONLY seeing success, fueling a hype train that doesn't stop. I thought I'd share a little bit of failure, because the lessons are still neat, and I'll carry them to my next experiment. My basic idea: A user uploads a graphic and additional context about the learning environment/audience. AI would draft a report of design errors, and then identify areas for improvement. It would also generate an “improved” version of the graphic and finally show a comparison shot and a discussion of the areas to be improved. Enter Google #Opal. It works by connecting nodes (input, generate, output) that all have Gemini capabilities to create an app-like experience. Unfortunately, I found that: 1. Google Opal does not encourage manual building. You’re able to customize nodes as you need, but I kept running into errors when doing the branching myself. I only had success when using the AI chat box to tell it what to build. 2. The AI builder is well-intentioned but works in small chunks. While it could generate the experience with interconnected nodes, any manual changes I made were not saved with the next wave of AI-powered edits. This made for an increasingly frustrating experience and squashed any chance of Opal being used for development. 3. Opal has access to Google’s leading models, but I struggled to build something truly usable. Ideally, I wanted to build a conversation where AI would ask users about their design choices, respond in real time, and then create a report based on that. That level of interactivity does not seem possible with Opal without extensive technical knowledge, meaning the sustainability of a product built on it would rely on expert developers. 4. Despite having access to the leading models, the output was arguably worse than what I would normally get with regular Gemini. Please take a moment to note the amended infographic. My primary concern is that it didn’t make minor changes to the design, it completely uprooted the foundations of the infographic to redesign in its own way. Opal is not a viable tool for development. It's still an “Experiment” by Google, so I'm silly for even trying! But I wanted to test it because I saw others share incredible on the surface Opal builds, though now that I think about it, without any follow-ups or evidence of the development process. As we share, build, and grow, I want to caution you on the dangers of hype. Do your own experiments. Keep questioning whether something works, works for you, or works for your entire ecosystem. There’s a difference, and that’s how you can derail the hype train. I might take my idea to another tool to see what’s possible. Google's AI Studio? A GPT perhaps? Who knows!

37

12 Comments

I don't know who on your ELT needs to hear this... But this "give 1st" partnership principle helped Ray Dalio (165th wealthiest person in the world) build a $125 billion company from 0. His quote... "When I had nothing, I was as generous as I could be with people who appreciated my generosity more than the higher levels of compensation others could afford to give them." My takeaway... If you want to influence partners who are bigger than you (or they have existing high performing partners) figure out ways to be strategically generous. Here's things I've done... When building partnerships with companies that were 10x to 100X larger than mine, I did these things: • Owned the build / ran surveys and research for them • Intro'd to key decision makers at key accounts • Owned / created ebooks based on joint-value • Sent reps personalized gifts (read Giftology book) • Sent referrals to their sales teams (without asking) • Organized joint-customer product feedback sessions • Enrich a leads list for them using your proprietary data • Market a webinar or report / help them hit their MQL goals Overall theme was how and I create value for us / them to demonstrate we will be a strong partner (able to drive results and easy to work with).

44

10 Comments

When advanced email attacks reach teacher and staff inboxes, it impacts instructional time, district operations, and community trust. Modern K-12 districts are shifting from reactive cleanup to proactive, AI-powered protection, without replacing their existing tools. Check out these resources form Abnormal AI to learn how districts are simplifying email security and protecting their communities. #EdTech #K12Cybersecurity #EmailProtection #StayReady #K12

14