The fact that Google I/O was immediately followed by Microsoft Build, with both conferences focused on AI, provided an interesting contrast between the two companies’ approaches to AI. Microsoft is interested in computers for AI integration, while Google is interested in smartphones.
As I sat in the Microsoft audience waiting for the keynote, there were no computers in sight, but when the keynote started, most of us pulled out our computers because those smartphones aren’t useful for writing. We don’t need one or the other. We need both.
Interestingly, the two technology companies that supported Microsoft’s announcement were not Intel or AMD but Qualcomm in PC and Nvidia in the cloud. This showed once again that if these two companies could put aside their animosity and work together more closely, they could probably secure their AI future. They are both on display as AMD and Intel prepare to compete and companies like Microsoft, Google, Facebook, OpenAI and Amazon ramp up their own hardware efforts.
Let’s talk about AI this week, and we’ll end with what I think is the best of the laptops released that use the new Qualcomm Snapdragon X Elite.
Computer or smartphone?
If you watched Google I/O, you saw a compelling demonstration of how your smartphone could be the most convenient way to interact with AI. Your smartphone is always with you, its camera is easy to use, and it can learn more about your personal life to create AIs that better match your unique needs.
Microsoft emphasized its AI in the computer that learns from your work habits, digital communications, gaming activities, professional interests, skills and abilities. The computer remains your work interface and the AI learns how you manage long-term communication to improve its understanding of your processes, skills and abilities.
Google’s approach should improve your personal life, while Microsoft’s approach should improve your professional life. However, you are not just one or the other; you are both. Therefore, for the AI to really get to know you and be more effective working and playing with you, it needs to engage with you professionally and personally.
You’ll want consistent AI across both platforms. Thus, Google and Microsoft are exposed to a future competitor that can bridge the two platforms or produce a new device that can be effective personally, like a smartphone, and professionally, like a computer.
Strangely, now, as I mentioned above about Qualcomm and Nvidia, the most beneficial approach for users is for Google and Microsoft to cooperate rather than compete, because their efforts complement each other.
Quality over performance
One of the things that worries me most about Microsoft’s approach compared to Google’s is the lack of attention to quality. If you have quality problems (and everyone has quality problems) and you increase speed without fixing those quality problems, you will make more mistakes. If these errors become widespread in AI systems, their frequency and impact will increase significantly.
The industry is tackling this problem backwards by focusing on increasing performance rather than quality issues. With people like Nvidia CEO Jensen Huang saying that there is no need to learn to code, the ability to find these bugs is decreasing as the number of errors increases as we increase speed. There is a real possibility that we will eventually encounter quality problems that accelerate faster than we can mitigate them. I’m already seeing a significant reduction in quality as companies develop AI.
The reason I focus on Microsoft here is not because it focuses less on quality than Google, but because Microsoft’s focus on development and creation makes the quality problem much more dangerous. Google’s evidence did not include coding. What its users created was more about entertainment, while Microsoft’s tools are used to build buildings, design vehicles and create new AI. Thus, the risks of this problem are greater in the work of Microsoft, which therefore has a higher quality requirement than that of Google based on the presentations of the two companies.
Advantages of AI
Microsoft presented a different and very compelling set of AI use cases. Whether it’s helping blind people by using AI’s ability to describe what it sees and providing instructions and descriptions based on that sight, or giving a paraplegic the ability to speak and move, it was a particularly compelling demonstration.
This could help teachers be more efficient while freeing up their time to focus more on individual students, and it could even help a father learn to play Minecraft with his young son. I found the focus on teachers particularly compelling because they are vital to ensuring future generations. Next was the training aspect of the tool, and the fun demonstration really caught my interest.
The latter was particularly interesting because I often lose interest in a new video game because, during the initial learning process, getting to grips with the unique aspects of the game, for me, is more work than fun. But I know I will enjoy the game more if I can get past this learning phase. In story-driven games, there are times when I just want to move forward with the story without spending a lot of time on any particular aspect of the game.
Additionally, it’s often more useful to have a trainer rather than an online guide to help you overcome the obstacle, and this is what Microsoft proved: an AI trainer that behaved and spoke like a human, that could see what you see and guide you. to a more successful and, for me, more fun gaming experience. This would take a lot of my anxiety out of buying a new game, with the result that I’ll be more likely to buy more games if they have this feature.
Office 365 widely adopts Copilot, allowing you to automatically create presentations from documents. It can be much more efficient to write a document while creating a presentation rather than jumping straight into PowerPoint.
I’ve often wanted to learn Adobe’s tools so I can use them to create better images and movies, but the learning curve is steep. With AI, the learning curve is greatly reduced, making it much more likely that I can become productive with the product immediately.
With AI, integrating even the most complex tools will become much easier, which could be huge for companies like Adobe that have powerful tools but whose sales are limited by customer fear of the associated learning curve.
Conclusion: Our approach to AI may be seriously flawed
I mentioned that we focus too much on speed and not on quality. I would add that this is where we need AI the most: decision support. This issue of quality versus speed is a case in point. For example, earlier this year, OpenAI’s board fired its CEO. It ended badly. Their decision was well-founded but not carefully considered. Instead of making OpenAI more secure, their failure arguably made the company less secure.
We often make decisions that, if we had thought about them carefully, we would have made differently, if at all. Looking back, we’re all smarter, but AI could help us avoid many of our critical mistakes.
In a way, this brings us back to quality. Rather than initially focusing on AI to make us more productive, shouldn’t we be focusing on helping ourselves create better, happier lives? Could this help us avoid the relationships, careers and mistakes that haunt us throughout our lives and ensure that where we are and where we end up is where we will enjoy the most and regret the least?
I think AI should focus on securing our best future, not just our most productive future, but that’s not the path we’re on right now.
Samsung Galaxy Book4 Edge
At the Microsoft event, Acer, Asus, Dell, HP, Lenovo, Microsoft (with Surface) and Samsung all showed laptops powered by the impressive new Qualcomm Snapdragon performance to run. Microsoft Copilot in full.
Most of the proposals were generic due to the speed with which these products had to be developed. Microsoft’s focus on Apple has resulted in the best Surface laptop yet. Acer had a great feature that turned on when Copilot was used, but Samsung did what it did with the old effort of Microsoft VR headsets and created the most powerful high-quality experience with a very clean design, the use of high-quality technologies and functions throughout, and by far the most beautiful OLED screen.
The Galaxy Book4 Edge is a Copilot+ computer powered by the Snapdragon X Elite processor. (Image credit: Samsung)
I had a hard time choosing which laptop to take: the Surface Notebook, which has long been a favorite, Acer’s lighting feature coupled with its exceptionally sleek design, or the wall-mounted effort offered by Samsung.
In the end, among remarkably similar offerings, the Samsung stood out as the one I’d buy if I needed a new computer. As a result, the Samsung Galaxy Book4 Edge is my product of the week.
Samsung is one of only two Windows PC vendors that can combine smartphones and PCs to create the consistent AI experience we want across both platforms. The other is Lenovo. Keep an eye on these two vendors in the future, as they may end up holding the future of AI-based personal technology. Oh, and hats off to Qualcomm for its Snapdragon X Elite that makes this possible.