Intel built a VERY fast and efficient AI research model

- Simplified AI -
Your gateway to discovering the world of Artificial Intelligence!

This week taught me why Claritin exists - I think I single-handedly kept Kleenex in business. Happy Spring!

A sneak peek of what’s below:

• What are Generative Adversarial Networks?

• Microsoft's AI tool makes scary good fake AI videos

• Are AI wearables worth the hype?

• Guess the real picture!

What are GANs?

Turning AI topics into a walk in the park, not a mind-boggling maze

Let's play a game! Further down in this newsletter, you'll find four images - I created three of them with AI, and one is a real photo. Can you figure out which one is real? (Answer at the bottom of the newsletter.)

This game is similar to the basis of Generative Adversarial Networks, or GANs for short, which were first introduced in 2014 by Ian Goodfellow. At the heart of GANs lies a creative dance of innovation and critique between two adversaries - The Generator and the Discriminator.

The Generator aims to produce synthetic data samples—images, text, etc.—that closely resemble real data, while the discriminator sharpens its discerning eye, learning to distinguish between genuine and counterfeit creations. As the training progresses, both networks engage in a relentless pursuit of improvement, with the generator producing increasingly convincing outputs, and the discriminator honing its ability to detect subtle flaws.

Pitting these two adversaries against each other allows each version to become better at what they do. Through this iterative process of competition and refinement, GANs excel at generating high-quality, diverse outputs across various domains, including image synthesis, text generation, music composition, etc. As the generator and discriminator engage in this creative duel, GANs continue to push the boundaries of artificial intelligence, unlocking new frontiers in generative modeling.

The implications of GANs extend far beyond mere imitation, unlocking the potential for AI to generate entirely new and original content. From realistic images of nonexistent landscapes to compelling prose that echoes the style of renowned authors, GANs are pushing the boundaries of creativity and innovation. This aspect will continue to be incredibly important in training AI models going forward, especially considering experts' concerns about the scarcity of high-quality training data.

Furthermore, GANs have practical applications in fields like drug discovery, where the ability to create realistic data samples is highly beneficial. In drug discovery, scientists need to find new medicines, which can be time-consuming and expensive. GANs assist by creating synthetic data that resembles real molecules and how they interact in the body. Through the utilization of GANs, scientists could expedite the discovery of treatments for various diseases, leading to faster and more cost-effective outcomes.

AI News Flash

Catch the latest AI news that is making all of the headlines this week!

• Intel's game-changing AI research system "Hala Point", a neuromorphic (human brain-like) research computer, can solve optimization problems 50x Faster, using 100x less energy

• With just one image and one minute of audio, Microsoft's Vasa-1 can create scary good AI avatars and deepfakes (Click the link and watch the videos, trust me)

• If you use Slack at work, you might be able to boost your productivity by summarizing your daily messages with its new AI Recap feature

• AI is revealing secrets about famous art pieces and snitching on the artists

• The US Military had a historic live dogfight between a human and AI-piloted fighter jets

AI Pictures of The Week

I made three of these Japanese Rock Garden pictures with AI, and one is a real photo. Can you guess which one?

AI Spotlight

Explore how AI is actively shaping our digital future

Are AI wearables here to stay? Or are they all hype?

Remember all of those futuristic TV shows and movies we watched growing up with robots, holograms, and artificial intelligence? Well, that future might not be here yet, but some companies are attempting to bring us closer to that vision by producing AI consumer devices.

The concept of AI Wearables, such as smart glasses and the Humane AI Pin, has been dominating tech news lately. However, early reviews are mixed. While some praise the innovation behind these products and ongoing efforts, many others are less impressed.

Popular YouTuber Marques Brownlee (known as MKBHD) called the Humane AI Pin "the worst product I've ever reviewed…for now," and if you search "Humane AI Pin" on Google, you'll see plenty of other scathing reviews flood your feed. To sum up the reviews: the Pin lacks functionality, performance is sluggish, and it falls short on battery life.

But this isn't just about one product. The real question is: do we even need AI-specific wearable devices?

As bullish as I am on a lot of technology, I see very little market potential in the future for AI-specific wearables. It's probable that we'll witness numerous startups attempting to produce the next hottest AI wearable product, only to ultimately fail in gaining traction.

The reality is that our phones can accomplish everything these wearables promise to do – and do it much better. Smartphones have evolved into indispensable tools in our daily lives, functioning as communication devices, cameras, navigation aids, entertainment hubs, and much more. With the vast array of apps available, smartphones offer a level of versatility and convenience that AI wearables will continue to struggle to match. Smartphones boast larger screens, more processing power, and longer battery life, rendering them far more practical for most users.

I don't see a scenario where a significant number of consumers would opt for one of these devices over their smartphones. Of course there are tech enthusiasts out there, but that smaller population just won't cut it. You would need a very large customer base to support building an AI hardware product and maintaining an AI model.

Technology adoption is always challenging, and someone has to be the pioneer. Just because these AI wearable devices aren't meeting expectations right now doesn't render them completely useless. Innovation thrives on iteration, and there's bound to be valuable lessons learned for future technology. Think back to the early iPhones compared to the latest models – the progress is undeniable. Did you ever imagine holding your first iPhone up to your face to automatically unlock it? Or tapping it on a scanner to quickly pay for your groceries? Probably not.

So ultimately, unless all of our smartphones stop working next week, AI wearables don't have a place in this world for the foreseeable future. But one thing's for sure: the journey toward integrating AI into our daily lives is well underway and I'm excited to see what happens next.