AI is seemingly everywhere, but ask most people to define it, and you will likely get blank stares. Honestly, we write about it every week, and we’re still not exactly sure how to explain it.

An article in The Conversation nailed it. AI, it turns out, goes far beyond chatbots.

AI systems have seven basic kinds of capability from recognition to interaction. If you’re eager to grasp what AI can really do, we think this article is well worth reading.

In this week’s edition, we’ve got another clever chatbot analyzer tool, AI for kids, AI’s flavor innovation, and an amazing ancient art discovery.

Let’s get started!

AI Chatbots

It's been almost a year since we launched this newsletter, and one thing is clear: not all AI chatbots are the same. Each has its strengths, quirks, and blind spots.

In recent months, we’ve featured Choosy Chat, which compares AI chatbots side by side, and NotDiamond, which selects the best AI model for specific inquiries.

Now, there’s OverallGPT, which goes a step further by not only offering AI model comparisons but also synthesizing all the responses to deliver key takeaways. This feature helps users save time and quickly grasp insights from multiple perspectives, making analysis more reliable.

We tested OverallGPT with a simple prompt: "What are the three most important skills for the future of work?"

The results included comparisons from OpenAI’s ChatGPT, Anthropic’s Claude 3.5 Sonnet, Google Gemini 1.5 Flash, and Meta Llama 3.1 (405B). What stood out was the synthesized answer:

• Adaptability and Continuous Learning: Flexibility and ongoing skill development are essential in an evolving work environment.

• Critical Thinking and Complex Problem-Solving: The ability to analyze information and solve complex issues is crucial for informed decision-making.

• Emotional Intelligence and Interpersonal Skills: Strong emotional intelligence and communication enhance collaboration and teamwork.

We were impressed when we tried again with a more complex question on how blockchain technology works and its applications beyond cryptocurrency. You can view the results here.

OverallGPT is fully transparent, presenting results exactly as generated by the AI. You can try it at no cost. When you sign up, you’ll get five daily credits.

AI for Kids

We’re happy to see a smart idea like this is already in the works: an AI chatbot that’s accessible and safe for kids.

Angel Kids is a new platform designed to create a secure, child-friendly browsing experience. Kids can explore, learn, and ask voice-activated questions in a controlled, age-appropriate environment.

Created by Nashville entrepreneur Tim Estes, alongside a team of AI experts and child development specialists, Angel Kids offers ad-free, data-secure exploration. According to the website, it avoids predatory monetization and ensures children’s data isn't collected or sold.

The platform will deliver personalized answers through curated multimedia content—images and videos—without requiring kids to navigate separate apps or websites. Using large language models (LLMs), natural language processing (NLP), and speech recognition, Angel Kids aims to provide tailored, child-safe responses. As the AI learns about a child’s interests, it will generate increasingly relevant and engaging answers.

Angel Kids features an exclusive Parent Portal, offering insights into a child's interests to encourage meaningful conversations. Customizable screen-time limits are also included, with gentle reminders when it's time to step away.

There’s no charge to join the Angel Kids’ waitlist and become a founding member before launch.

AI Food Science

You might not think much about how your favorite snacks or beverages get their flavor—but new AI-powered research could revolutionize how quickly and precisely those flavors are created.

Scientists at the University of Southern California (USC) and the University of California, Davis have developed FOODPUZZLE, an AI tool that speeds up flavor creation. With a dataset of nearly 1,000 foods and 1,766 flavor molecules, this AI can predict and complete molecular flavor profiles more accurately than ever before.

Faster, smarter flavor development means new, more diverse products hitting shelves quickly. Think better plant-based foods, healthier snacks, and even personalized flavors to tickle your tastebuds.

How it works: The AI analyzes the molecular makeup of foods, identifying flavor molecules that define taste. By predicting missing molecules or combining existing ones, the AI mimics how flavors interact chemically, allowing scientists to create precise flavor profiles more efficiently than traditional methods.

By combining AI predictions with lab testing and human taste panels, this research could produce food products tailored to consumer preferences, while cutting down on trial-and-error methods in food science.

This AI-driven approach to flavor innovation could not only improve the taste and variety of your food but also accelerate the arrival of healthier, more sustainable options. The future of flavor is smarter, and it’s coming to a grocery aisle near you.

AI & History

AI may hold the key to the future, but it’s also unlocking secrets of the past. Recent advances in AI are helping to uncover ancient wonders hidden for centuries.

In Peru’s Nazca Desert, scientists deployed AI to discover over 300 previously unknown geoglyphs—doubling the number of known Nazca Lines in just six months of fieldwork.

For those unfamiliar, the Nazca Lines are massive, human-made figures etched into the earth by the ancient Nazca civilization between 200 BC and 500 AD.

Their vast scale and intricate designs have long fueled speculation that extraterrestrials were behind their creation, as many figures—some stretching hundreds of feet across—are only visible from the air. These discoveries provide fresh insights into the Nazca culture while demonstrating AI's transformative power in archaeology.

Before AI, it took nearly a century to uncover 430 geoglyphs. Now, with satellite imagery and AI, researchers have dramatically accelerated the process, identifying figures that remained hidden for millennia.

• The AI model, developed by researchers at Yamagata University in Japan, was trained to detect the faint, eroded lines of the geoglyphs, working 20 times faster than human-led searches.

•  Field expeditions undertaken between September 2022 and February 2023 confirmed 303 of the AI-suggested geoglyphs.

• The newly discovered geoglyphs include abstract animals, severed heads, potential ritual scenes, and even a 72-foot-long killer whale holding a knife.

You can view some of the newly discovered geoglyphs here.

Looking ahead, researchers expect AI to uncover another 250 geoglyphs in the coming years, deepening our understanding of the Nazca civilization and their mysterious earthworks.

AI in the News (in case you missed it)

• Tech Giants Bet Billions on AI Agents That May Reshape Commerce as We Know It. Read here.

• Happy, sad or angry? AI can detect emotions in text. Read here.

• Google’s search engine’s latest AI injection will answer voiced questions about video and photos. Read here.

• Refik Anadol Is Launching the World’s First Museum of A.I. Art. Read here.

Did you receive this from a friend, colleague, or family member? Please subscribe.

Have you tried a new AI-powered product? Let us know what you think.

Know anyone who might like our newsletter? Recommend it, and help grow our community!

Here.Now.AI Editorial team: Lori, Justin, and Lisa