I watched IBM Think last week. As I listened to IBM CEO Arvind Krishna talk about how IBM is willing to configure hardware, software, and services for every customer, it struck me that the one thing we are not doing with AI is thinking it through. We are throwing AI at everything and not prioritizing where we need it to focus.
We had the same issue with nuclear energy early on. We could focus it on safe, low-cost energy generation or on destroying the planet. We initially focused on destroying the planet, followed by unsafe energy generation (Chernobyl, Three Mile Island).
A meme on social media says, “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.”
Let’s explore this lack of thinking about AI use in the context of last week’s announcement at Computex of the impressive Intel Lunar Lake platform.
We’ll close with my Product of the Week, “Trustworthy,” which exemplifies a better direction for AI and will help us better prepare for the future problems that all our families might face.
IBM Think
The Think slogan goes back to Thomas Watson, Sr. In 1915, as a sales manager for NCR, he developed the saying after becoming frustrated that no one on his executive team seemed to have any ideas. He argued, “The trouble with every one of us is that we don’t think enough. We don’t get paid for working with our feet — we get paid for working with our heads.”
This lack of thinking is what we’re not addressing when it comes to AI.
Like any powerful tool, AI is neither good nor bad. Like nuclear technology, it depends on how the technology is used. Currently, there is a lot of focus on building ever more powerful AI weapons, some of which are pretty scary. This drone-delivered weaponized Chinese robotic dog could be a scene from “The Terminator,” and that is far from the only example.
Back at the Beginning and IBM Watson
When IBM started Watson, it was accidentally focused on winning game shows to demonstrate its capabilities. However, much of the initial work was on providing better and more timely medical advice.
While IBM’s inability to reliably access medical data and the potential liability of getting that wrong forced it to pivot away from this market, it did highlight that IBM was following the “Think” process because better medical care should be an exceedingly high, if not the highest, priority. It should come way before AI weapons systems.
Medical costs, particularly here in the U.S., are incredibly high, while medical quality is largely deficient, which seems like the worst of all worlds. AI could address the cost and the quality of medical services, but right now, that does not appear to be a top priority, giving way to the more troubling uses of AI.
I think rather than focusing on throwing AI at everything, we should have a priority list of things we want AI to do. Toward the top would be a far better digital assistant.
Lunar Lake vs. Snapdragon X Elite
Your digital assistant will need to be with you while you work, and much of our initial AI use will be tied to making us more productive. However, we need to push up the quality requirements of this effort because, in the last head-to-head comparison, Microsoft Copilot ranked last. It is interesting to note that the current iteration of ChatGPT, which is the basis of Copilot, ranked second while Perplexity ranked first. We don’t only need AI with us; we need the information it provides to be accurate and useful, and right now, Copilot is falling short.
Copilot+, which includes Recall and Cocreator, was designed to work on Qualcomm’s Snapdragon X Elite processor, and initially, both AMD and Intel were bypassed. However, both companies have put in a ton of work to close this gap, and it looks like the next big AI chip drop will come from Intel and its high-end client processors code-named Lunar Lake.
Lunar Lake represents a massive effort from Intel and brings to market a level of capability we weren’t expecting until 2026. Still, it will also be more expensive, allowing Qualcomm to likely own mainstream notebooks until AMD’s more affordable alternative comes to market.
However, and this again goes to priorities, you’d think workstations and desktop PCs would have been prioritized because that is where professional creators still live, and even workstation laptops aren’t included in the initial Copilot+ efforts. The closest thing to a Copilot+ workstation is the Qualcomm Snapdragon Dev Kit for Windows which people who want this AI functionality in a desktop are buying up.
Given that desktops and workstations don’t require the power of laptops and are used by developers, who are more Microsoft’s primary customers than end users, you’d think desktop and laptop workstations would be a priority. Nvidia prevailed on Microsoft to embrace their platform, and AMD is right there with them, but with a gaming focus.
Focusing on gaming is important because smart non-player characters (NPCs) could make an enormous difference in games, both as opponents and partners. Imagine, for instance, having a bonded pet like a dragon that could fight and protect you intelligently, let alone an in-game helper who could guide you through some of the more challenging puzzles (I hate puzzles in games).
Wrapping Up: AI Priorities
Back to priorities: I think AI should first help us make better decisions and then focus on those things that take the pleasure out of life, like cleaning, doing the laundry, driving in traffic, checking our work (but not doing it for us), and helping us form better relationships or make relationships work.
In the end, AI could make the world a better place or one that was much, much worse. Personally, I’d prefer the former over the latter. I don’t need more difficulty in my life, but I could certainly use more help.
Trustworthy: The Family Operating System
On the topic of where I think AI should be focused, it is helping us manage our personal and family life.
From maintaining documentation so that, should something happen to us, those left behind can more quickly and easily manage our estates and bills, to keeping track of critical records like mortgages, birth certificates, credit cards, retirement accounts, insurance, and passwords, the Trustworthy Family Operating System helps with the annoying tasks that can take the fun out of your personal life.
Trustworthy uses AI, but it is memoryless, as it doesn’t learn from customer data. It caught my attention because I’ve been dealing with a lot of this stuff lately and really could use some help like what Trustworthy provides. This video provides an overview.
Given my focus on AI for good, Trustworthy seemed like an ideal example of where we should focus AI. In this case, it better prepares us for disruptive events, recovering from natural disasters, death, getting ready for retirement or relocation, and just bringing a bit of control and peace of mind to our overly complex daily lives.
Trustworthy starts with a modest service for protecting financial information for $120 a year. It jumps up to $240 a year for families that also have business interests, then $590 with three hours of service to help you throughout the year and to initially get set up, and finally, $990 for those needing a lot of help, such as an aging relative who needs six hours of help to get decades of mess into order.
Data is encrypted. It uses multi-factor authentication (biometrics) to help ensure this critical information stays safe and uses tokens to obfuscate sensitive information further. It also regularly redacts sensitive information like driver’s license numbers to protect it from prying eyes.
I think Trustworthy is closer to where we need AI-focused: making our lives easier, eliminating stress, and keeping us safer rather than taking our jobs, eliminating our incomes, or powering deadly weapons. As a result, Trustworthy: The Family Operating System is my Product of the Week.
Read the full article here