In the rapidly evolving landscape of artificial intelligence (AI) and quantum computing, the opinions of industry leaders can significantly influence the direction of technological advancements. Yann LeCun, Meta’s chief AI scientist, recently offered a grounded perspective on these technologies, providing a contrast to the often hyperbolic narratives surrounding AI’s future capabilities and the potential of quantum computing.

AI’s Journey to Sentience: A Long Road Ahead

LeCun, a pioneer in deep learning, expressed skepticism about the imminent arrival of artificial general intelligence (AGI) – AI with human-level intelligence. Speaking at the Viva Tech conference in Paris, he highlighted the limitations of current AI systems, which, despite their ability to process vast amounts of text, lack the common sense necessary for true sentience. This view contrasts with Nvidia CEO Jensen Huang’s assertion that AI will rival human intelligence in less than five years, as reported by CNBC. LeCun’s stance reflects a more cautious and realistic assessment of AI’s current trajectory.

The Hype Around AGI and Quantum Computing

The pursuit of AGI has driven significant investment in AI research, particularly in language models and text data processing. However, LeCun points out that text is a “very poor source of information” for training AI systems to understand basic concepts about the world. He suggests that achieving even “cat-level” or “dog-level” AI is more likely in the near term than human-level AI. This perspective aligns with the broader consensus in the AI community that AGI remains a distant goal.

Multimodal AI: The Next Frontier

Meta’s research into multimodal AI systems, which combine text, audio, image, and video data, represents a significant step forward in AI development. These systems could potentially uncover hidden correlations between different types of data, leading to more advanced AI capabilities. For instance, Meta’s Project Aria augmented reality glasses, which blend digital graphics with the real world, demonstrate the potential of AI to enhance human experiences, such as teaching tennis techniques.

The Role of Hardware in AI’s Future

Nvidia’s graphics processing units (GPUs) have been instrumental in training large language models like Meta’s Llama AI software. As AI research progresses, the demand for more sophisticated hardware will likely increase. LeCun anticipates the emergence of new chips specifically designed for deep learning, moving beyond traditional GPUs. This shift could open up new possibilities in AI hardware development, potentially leading to more efficient and powerful AI systems.

Quantum Computing: Fascinating but Uncertain

LeCun also expressed doubts about the practical relevance of quantum computing, a field that has seen significant investment from tech giants like Microsoft, IBM, and Google. While quantum computing holds promise for certain applications, such as drug discovery, LeCun believes that many problems can be more efficiently solved with classical computers. This skepticism is shared by Meta senior fellow Mike Schroepfer, who views quantum technology as having a long time horizon before becoming practically useful.

A Balanced View on Technological Progress

LeCun’s views offer a balanced perspective on the future of AI and quantum computing, tempering the excitement with a realistic assessment of current capabilities and challenges. As the tech industry continues to explore these fields, it’s essential to maintain a critical eye on the practical implications and timelines of these technologies. The journey towards more advanced AI and the realization of quantum computing’s potential will likely be a long and complex one, requiring sustained effort and innovation.

In conclusion, while the future of AI and quantum computing is undoubtedly exciting, it’s important to approach these fields with a realistic understanding of their current state and potential. As LeCun’s insights suggest, the path to AGI and practical quantum computing is longer and more nuanced than some of the more optimistic predictions imply. The tech industry must continue to push the boundaries of what’s possible while remaining grounded in the realities of technological development.

In the realm of visual technology, the quest for more immersive and realistic experiences never ceases. Holography, the science of creating three-dimensional images, has long been a subject of fascination and research. Now, a groundbreaking study led by Professor Tomoyoshi Shimobaba of the Graduate School of Engineering at Chiba University has introduced a novel deep-learning method that simplifies the creation of holograms. This innovation allows 3D images to be generated directly from 2D photos captured with standard cameras, marking a significant advancement in holographic technology.

Traditional holography involves capturing an object’s three-dimensional data and its interactions with light. This process demands high computational power and specialized cameras for capturing 3D images. This complexity has restricted the widespread adoption of holograms, despite their potential applications in various sectors like medical imaging, manufacturing, and virtual reality.

Deep learning has been making waves in the technology sector, and its application in holography is no exception. Previous methods have employed deep learning to create holograms directly from 3D data captured using RGB-D cameras, which capture both color and depth information of an object. This approach has circumvented many computational challenges associated with traditional holography.

The team from Chiba University proposes a different approach based on deep learning that further streamlines hologram generation. Their method employs a sequence of three deep neural networks to transform a regular 2D color image into data that can be used to display a 3D scene or object as a hologram. The first neural network predicts the associated depth map from the color image, providing information about the 3D structure of the image. The second and third neural networks are responsible for generating and refining the hologram, respectively.

One of the most striking aspects of this new method is its speed. The researchers found that their approach outperforms current high-end graphics processing units in speed. Moreover, the method is cost-effective as it doesn’t require expensive equipment like RGB-D cameras after the training phase.

The implications of this research are far-reaching. In the automotive industry, for instance, this technology could revolutionize in-vehicle holographic systems, presenting necessary information to passengers in 3D. The U.S. Department of Transportation has been exploring the potential of such advanced display technologies for enhancing road safety. Additionally, the technology could find applications in high-fidelity 3D displays, heads-up, and head-mounted displays, further augmenting the development of ubiquitous holographic technology.

The introduction of deep learning into the field of holography has the potential to solve many of the challenges that have hindered its widespread adoption. By simplifying the process and making it more cost-effective, this new method could pave the way for holography to become a more integral part of our daily lives, from healthcare to transportation and beyond.

The research, titled “Multi-depth hologram generation from two-dimensional images by deep learning,” was recently published in the journal Optics and Lasers in Engineering.

Housemarque has a new video game that struck me as so unique and fascinating that I was compelled to buy it.

I made the decision to buy the game without watching a ton of demos or reading up on internet reviews.

It was a no brainer. From the trailer, I could already tell that Returnal was similar to many other games. It was about a strong, smart, and capable protagonist who had to conquer the odds. That is not what compelled me to make the purchase.

The main character Selene is a woman. That in itself is an anomaly. But what makes her even more of an oddity is her age. Selene is a middle-aged woman.

Female videogame characters are typically youthful and sexy. The few who are not young are elderly women revered for their wisdom or hated for their villainy.

So a middle-aged female lead captured my attention so strongly that I just had to buy it.

I could finally see someone like myself in a game. You see, I am a 50-year-old woman who has recently found out that (surprise!) middle age is not exactly what I thought it would be.

At 50, I do not feel young. I do not feel old either. My body is not as fast as it was in my youth, even though I am in good shape.

In fact, I feel like I am at my best ever. But there is a certain invisibility that middle aged women can all identify with.

People can be rather awkward around a 50-year-old woman. Rarely do they start conversations. When they do talk to you, they ask about your children.

This adds another layer of awkwardness and invisibility for me as a childless woman.

With no children, they cannot imagine anything else to talk about.

When you find a woman in a game, she could be a companion or just there to add color to a party. She is not the star.

We can all remember how big of a shock it was for fans of Metroid to discover that the hero they had admired for so long was really a woman.

It was supposed to be a shocking plot twist. And it did shock many.

Today, gamers who want to play as women have many more opportunities than ever before. Some games even allow you to create the character you want to play.

But these female characters are almost invariably young.

To be honest, there are advantages to playing younger characters. Younger bodies are stronger and more agile. So it makes sense to have a main character in their 20s or 30s.

In Tomb Raider’s 2013 reboot, Lara Croft’s youth and experience are a huge part of what makes the story powerfully inspiring.

She was a fresh college graduate who found herself alone and injured after surviving a shipwreck.

Her inexperience meant that she had to challenge herself in ways she had never done before. She was discovering her own strength and we were there to watch her develop resilience. It was an awesome journey.

As a 50-year-old gamer, I cannot help but imagine what Lara Croft would be like two or three decades older. What challenges would she face? How would she tackle them? What allies and enemies would she have acquired along the way?

We have no way of knowing. Because the creators are not particularly interested in Lara Croft post 30.

At this point, you are probably wondering just how many gamers are wondering about the same thing I am wondering about.

I looked up the numbers, and female gamers my age are almost as many as male gamers. There is only a 2% difference in the numbers.

My point is that if we can have such a variety of middle-aged male characters, then women my age deserve representation, too.

I would love to see more female gaming characters who are my age saving the world, solving mysteries, and slaying the occasional dragon.

Speaking of dragons, I can remember when the dragons in video games looked more like ducks than dragons.

Not too long ago, Ubisoft argued that they could not increase the playable female assassins in Assassins Creed because they were difficult to animate.

We have made some progress since then, thankfully. But we still have a long way to go.

I love that Selene features on the cover and stars in the game. I feel validated by her presence.

Selene represents the best aspects of being middle aged. Young people are more ambitious but also vulnerable to people-pleasing.

Selene comes armed with life experience the way only a middle-aged woman can be.

She is sure of herself and understands who she is and what she stands for. She is guided by her own internal compass and not other people’s opinions.

She has the confidence to steal a spaceship and go on a mysterious quest.

She is decisive and confident in her ability to handle whatever comes her way. She goes after what she wants without apology. I would love to see more characters like Selene.