SUBSCRIBE NOW!

Vol. 43 No. 8 — October 2023

EDTECH
Trying Out New Technology and AI Tools
by Brian Pichman

Turning our attention to 3D printing and VR/AR, we observe similar patterns of evolution. Around the mid-2000s, 3D printing gained traction as patents expired, allowing enthusiasts to create more affordable objects during the hacker and maker movements. By 2010, the race for superior desktop 3D printers began, even earning a mention in the 2013 State of the Union address. Sadly, some loved 3D printing companies fizzled out, and the support for the products along with it, because of market saturation while adoption rates slowed.

A dive into two of my favorite technologies—VR and AR—provided similar stories. If we count Google Glass in 2013, a futuristic vision emerged in which everyone would be wearing some sort of stylish headset display. Three years after, HTC VIVE and Oculus Rift (now a Meta offering under a different name) entered the scene followed by Microsoft when it took its shot with the HoloLens, offering both VR and AR experiences. Despite hailing from billion-dollar corporations, Google Glass and Microsoft HoloLens ultimately exited the market. Personally, I’m very curious about the adoption rate of Apple’s headset at its price point: $3,499 at release. Even as a loyal Apple fan, it’s impossible for me to justify getting one.

This all brings up several questions: How do we stay on the correct side of the curve? How can we anticipate a product’s lifecycle, so we suddenly don’t lose support for it? More importantly, which type of technology will push through hype cycles?

Even AI faces its own hype cycle, as discussions shift toward ethical concerns and data accuracy. There are great resources available to help us understand the hype cycle of technology. A fine place to start is Gartner Hype Cycle, which monitors emerging trends and plots them along a bell curve (gartner.com/en/research/methodologies/gartner-hype-cycle and en.wikipedia.org/wiki/Gartner_hype_cycle).

How to Evaluate Technology

So, how do we stay on the right curve, and how do we choose what product to adopt? It’s like adopting a new furry pet friend. Will this new kitty be nice and gentle and want cuddles all the time? Or will it quickly become the alpha and even boss around the dog in the house? (The latter is what happened with my cat, but I love her all the same.) Not every piece of technology you gravitate toward will be the one that remains standing, but it can play an important role as you explore that innovation trigger. Here are some easy ways to evaluate the longevity and flexibility of a product:

• Longevity—Avoid products with a projected life span longer than the rapid pace of technology evolution. Is the product supposed to last 5-plus years to reach ROI? If so, I would suggest avoiding that product because technology evolves faster than that, and it will be obsolete quickly.

• Transferability—Assess the potential to migrate data from a product, ensuring seamless transitions. Is this product transferable? For example, let’s say it is an AI tool for which you must train and correct the language models. If its parent company closes its doors, can you take what you built and export it (usually in a JSON format) so that you can import it into another tool?

• Company history—Evaluate the stability of the company behind the technology and its track record. How long has the company been in business? Does the CEO have a history of spinning up short-lived businesses? Now, that’s not to say a brand-new company will fail, but it’s something to consider in order to understand the level of risks involved.

• Community engagement—Gauge the willingness of staff to embrace and support the technology. Do you have staffers that are willing to learn the product and help shepherd its success with the community? You can’t buy the cutest pet at the pet store if the family won’t embrace it. The same goes with new technology.

Focusing on AI Tools

With that being said, let’s dive into AI as a whole and break things down as to how people are using it. The challenge with all AI models is garbage data in, garbage data out. If you noticed, ChatGPT has been a little less accurate as its usage has grown—such as getting simple math problems incorrect or making up six fake cases for which a federal judge fined two lawyers and a law firm $5,000 for using ChatGPT to research court filings. The concern is that AI models might encounter challenges as they are exposed to a broader range of queries. But in very structured approaches, with training data on top, OpenAI can be a strong conversational tool. AI presents boundless potential across various applications. While AI models hinge on data quality, AI tools exhibit diverse capabilities, such as the following:

• Helping build out Excel formulas: Formula Dog (formula.dog)

• Creating automated social media posts: Ocoya (ocoya.com) or Missinglettr (missinglettr.com), which can take your data off your website and turn it into viral posts

• Creating audio from text: Fliki (fliki.ai) or Amazon Polly (aws.amazon.com/polly)

• Creating video from text that pulls from stocked videos: Pictory (pictory.ai)

• Creating animations from text: SteveAI (steve.ai)

• Creating automated welcome video messages with custom text using your own voice and face—such as with BHuman (bhuman.ai) and ReachOut.AI (reachout.ai)—or using a realistic human-generated voice with either AI actors or prerecorded live actors

• Tools such as Chatbase (chatbase.co) allow you to build your own learned data from documents to have a conversational AI provide support on what you teach the model.

• MeetGeek (meetgeek.ai) is my favorite notetaker AI assistant. It joins all of your virtual meetings, transcribes the interaction, and sends you summaries and action lists.

• If you want to get better at speaking in meetings or presentations, take a look at Poised (poised.com), an AI-powered live coach that measures your performance over time and even helps you prep for meetings by building out agendas and talking points.

I’ve listed all of these and more at links.evolveproject.org/aitools. You can also play with many of these tools at links.evolveproject.org/aidemo, with a free sign-up for 60 days.

Products in the AI Education Space

Within the STEM technology space, there have been a lot of products over the years that have incorporated the ability to experiment and program with AI as part of their core functionality. Similar to the previous conversation about how to measure a company’s viability, when looking at STEM technology, you want to pick products that as a learner’s skill set grows, so can the complexity of the activities of the product. In other words, you can let a 3rd grader program the same robot with the same difficulty level as a seasoned coder in high school—just using increasingly complex programming methods (such as moving from block-based programming to programming via Python language). One of my favorite companies that has been doing it the longest within the robotic education space is BirdBrain Technologies.

BirdBrain Technologies has been a pioneer within the robotics education space. Founded at Carnegie Mellon University in 2010, BirdBrain leverages a lot of research-based learning styles and activities that help promote gender equality and diversity, with a focus on computer science, engineering, math, and robotics. Its Finch 2.0 is a cordless Bluetooth-connected robot in the shape of a bird. You can program a variety of sensors and outputs from distance sensors, light sensors, buzzers, lights, motors, and more. But the real magic begins when you program through its easy-to-use application that allows for some AI learning. Here is a great link to get you started: birdbrain technologies.com/blog/all-about-artificial-intelligence.

As with all of its other products, BirdBrain’s robots are designed to work with a wide range of ages and experience levels. You can code the robots with icon-based apps for little kids, and there are blocks-based environments for kids ages 9 and older and programming languages such as Python and Java for older teens.

If you’re more into how the Tesla Model S and DJI’s drones work, there is a company that caters to your technology favorites. Robolink has a variety of awesome products that explore AI coding and intelligent robots. For example, its cutely named Zumi is a trainable self-driving car. You are walked through simple activities that get progressively more challenging until you get Zumi to drive and navigate all on its own. If you are desperate for a flying vehicle, you can leverage Robolink’s CoDrone, teaching it to land on specific color squares, avoid obstacles, and more. Both of these products allow you to take early learners who just understand block-based coding and transform them into full-blown Python coders.

Kai’s Education stands out as a pioneer in the realm of education technology, boldly integrating AR, VR, and AI into a single transformative platform. Through its innovative products such as Kai’s Clan and KaiBot, this company is redefining the education landscape by offering an immersive VR experience designed to engage and inspire young coders. Kai’s Clan takes collaboration to a new level by enabling participants to code robots that navigate a physical play mat. But what truly sets this experience apart is the integration of AR/VR/Internet of Things technology, allowing learners to orchestrate the robots’ interactions with one another. KaiBot is a trailblazing hybrid robot that shatters barriers with its inclusive approach to coding education. Utilizing screen-free coding cards as scaffolding, KaiBot guides learners from coding cards to Blockly and even Python, all while having them be able to debug code. Its provision of Braille-encoded cards opens coding’s doors to visually impaired individuals, expanding the joy of learning to a broader audience. Across all of the products is the ability to translate block-based code into Python commands. This encourages experimentation and enables learners to delve deeper into coding, fostering a sense of ownership over their creations.

On the heels of AR/VR, I want to share a great product from PlayShifu. Among its wide variety of educational games, Tacto gives a new tactile play experience by transforming tablets into interactive gaming boards. With game sets that teach coding and provide an interactive doctor’s kit, Tacto is a complete play package that works by placing figurines on the tablet screen.

Other products from PlayShifu are Orboot’s educational globes (Earth, Dinos, and Mars). Each globe provides AR experiences by scanning it through the Orboot app using any iOS or Android device. You can explore cultures and famous locations on Earth, check out space missions on Mars, or learn about your favorite dinosaurs. A variety of self-paced learning activities allow a group of people to surround a tablet and the globe and get an immersive learning experience.

I know I’ve talked a lot about movable robots and have said nothing about reading. So, if you’re more of a bookworm than a robot geek like me, let me introduce you to Luka. Luka is a smart product that can read books to you. Simply hold up a book that’s within its library, and Luka will read that story to you, along with cute eye animations to keep the engagement level. If there is a book that’s not in its library, you can read the book to Luka, and it will learn it so it can read it back to you.

Lastly, I want to talk about ARpedia, an amazing company that recently won a variety of awards for its books that are combined with interactive AR experiences. With an extensive library covering diverse educational topics such as animals, space, and dinosaurs, these books offer an immersive reading journey enhanced by interactive components displayed on a tablet screen. Imagine chasing dinosaurs through cutouts and watching them come to life on the tablet or coloring your astronaut and going on a space mission around the solar system. There is endless play potential with ARpedia, making reading a fun experience. In addition, there are many other products in the market space, such as technology that allows AI to be used as part of our toolkit and products that teach us how to program with AI.

Whatever your journey, I would encourage calculated risks to explore new technology—call it a pilot or a beta program. It would be a better use case to be on the cusp of trying out new technology instead of sitting on the sidelines to see what others end up doing with it. You can leverage the same approach I explained earlier with evaluating options. Technology shifts faster than it ever has before. It can have massive adoption rates, but also experience massive drop rates (Meta’s Threads, for example). But if you look at usage graphs, it does follow a similar hype cycle that Gartner outlines.

One last piece about AI: It’s a tool that is only as good as the data and the model it sits on. As with all technology, it can be used for really, really good things. I encourage libraries and other educational institutions to explore how to make it a tool that people can use.

The Five Stages of New Technology Adoption

According to Gartner, every kind of technology goes through five stages. Here’s my take on the various stages.

• Innovation trigger—The journey begins with the spark of a potential technological breakthrough. Initial stories of proof of concept capture the attention of media, sparking a wave of excitement. Although functional products may be scarce and the feasibility for commercial success remains uncertain, the anticipation is palpable.
• Peak of inflated expectations—The initial wave of excitement leads to a surge in success stories, often intertwined with a setback or discovered limitations. Some companies take bold strides forward, while others remain hesitant. The technology becomes dotted with instances of triumph and failure.
• Trough of disillusionment—The fervor begins to wane, as experiments and real-world applications fall short of the grand promises. Companies producing the technology undergo a process of refinement or dissolution. Investments persist only if the survivors manage to enhance their offerings to meet the expectations of early adopters.
• Slope of enlightenment—Amid the disillusionment, a clearer picture emerges regarding the tangible benefits the technology can bring to the world. Instances of success start to crystallize, spreading a deeper understanding. Second-generation and third-generation products emerge from the technology providers. Enterprises, with growing confidence, allocate resources to pilot programs.
• Plateau of productivity—The point of mainstream adoption starts. Definitive criteria for assessing the viability of providers come into focus. The technology’s broad applicability within the market begins to manifest its worth. The investment of time and resources now pays off substantially, as the technology establishes its relevance across various industries.