Where AI is having the greatest success at the moment is with unstructured image-based data where, whether we are talking facial recognition or medical imaging, it significantly speeds up accurate identification of the image.

But AI can and does go farther with medicine in that it’s also used to identify the cause of a related illness and recommend the most efficient way to cure or mitigate that illness. If you’ve watched science fiction TV shows and movies, you’ve seen medical scanners that can better identify illnesses and injuries automatically. NVIDIA’s AI technology is on that critical path. This makes the related process more efficient and accurate and moves the timeline for creating these more advanced systems ahead significantly.

The Importance of Medical Imaging

Medical imaging is one of the most important tools in modern medicine today.

There is two-dimensional imaging for screening and early detection. Three-dimensional imaging layers on special understanding and quantitative measurement and segmentation. The fourth dimension adds temporal information, such as illness progression, that is essential for diagnosing and planning treatments. If an illness is progressing quickly, the responses need to be more invasive and higher risk, while a slowly moving or static progression may result in no medical response other than regular future observation to assure it doesn’t start spreading faster.

According to NVIDIA, using medical images with real-time, deep learning AIs and computer vision brings the industry into a fifth dimension where practitioners can get a holistic view of the patient, navigate within the human body to look for causes, and get a far stronger sense of the damage being done by the disease.

They can then use this information to plan actions while tracking progress and changes to the disease during the process. This, in turn, can help surgeons plan related procedures and surgical tasks.

NVIDIA’s Impact on Imaging

NVIDIA is aggressively operating in the medical segment and partnering with companies — like United Imaging, Fujifilm, Philips, Canon, Accuracy, and others — to provide the computational infrastructure needed to implement a comprehensive solution designed to improve image quality, lower radiation dose for X-Ray implementations, and run the related AI application to assist with the resulting diagnosis.

Much of the innovation of late has come from advanced sensors at the device level, increasing doctors’ ability to identify common and not-so-common problems and illnesses. But as sensors advance, they capture more data, and the related AI back end has to evolve to both absorb that data and provide the deeper insights that this additional data enables.

An example of this is Siemens’ Naeotom Alpha, a photon-counting CT scan, that improves image acquisition and reconstruction by reducing electronic noise while lowering radiation doses. This last is important, because CT scanners can significantly increase cancer risk due to the amount of radiation they use. Another example is Advanced Breast-CT’s nu:view product that uses spiral CT combined with photon-counting technology to deliver a compression-less breast exam that is more accurate.

NVIDIA provides innovators with the computational foundation to create the next generation of AI-backed, enterprise-class imaging platforms.

Improving Care

Medical events are often misdiagnosed or remain undiagnosed because of the quality of the imaging and the knowledge of the doctors, putting life and continued health at increased risk as people age.

These AI-based imaging advancements are helping assure that this imaging and diagnosis problem will improve sharply over the rest of the decade and hopefully increase our chances of as long and healthy a life as possible.

As a result, this NVIDIA AI effort is already changing care in an increasing number of health care institutions.

As you would expect, there has been a huge shift from working in the office to folks working from home. There was also a push to get people to go back to the office. But this report doesn’t reflect any significant success in that regard and suggests that the last six months or so have been relatively static regarding those coming in and those remaining remote.

As we look at office utilization, however, the report shows a massive shift between individual offices and the collaboration spaces that were once connected to them, both of which went into sharp decline, whereas general meeting spaces and casual social spaces doubled and quadrupled.

This suggests that office buildings configured for work prior to 2020 and those configured for work today need to be very different. Office provisioning needs to be both more flexible long term and more automated to address what is likely to be a moving target based on worker availability, changing needs, changing technology, and changing priorities across the segment.

Let’s talk about this survey and what it means for the future of the office:

The Deep Need for In-Person Social

The report appears to suggest that as we moved from coming into the office to working remotely, people became disenfranchised both from their companies and from their co-workers and want to go back into the office — not necessarily to work, but to reconnect with the band they operate under and the people they work with and for.

At least for now, this social requirement is driving office use and suggests far less needs to be focused on individual working spaces and far more on group meetings and particularly social interaction. It is interesting that the dedicated collaboration spaces that were part of the old in-office work model have declined by 50%, according to this study. But that was because those spaces were designed for people who were in cubicles or offices and occasionally needed to work together, not people who were remote and largely feeling the need for a deeper connection. I think that’s why the collaboration spaces declined by half, while the social areas nearly quadrupled.

Companies appear to have figured out largely how to collaborate well remotely, but the social interaction need wasn’t being met, so they are using the office space to close that gap. This suggests that a lot of the companies currently building collaboration spaces will find those spaces won’t be well utilized, because employees are coming in to socialize more than collaborate. The tools and space they want should reflect that need, and collaboration rooms don’t meet it.

Office of the Future

The difference in verticals in the study shows that government employees are largely staying home and not coming in much at all, while flex workers, those paying for and sharing offices, are coming in two to five times the rate of government employees. This may reflect that government work is largely non-collaborative and largely bureaucratic.

The office of the future not only needs to address the vertical industry and unique needs of the company today, but it needs to be flexible to address the needs of tomorrow to retain staff and drive a strategic agenda. This suggests future offices need to be flexible first and foremost, allowing for easy and rapid reconfiguration, so the changing needs of the employees are rapidly identified and better met. This last suggests far higher levels of employee monitoring and reporting, so operations and space planning can see trends and better anticipate and respond to employee needs with respect to where and how they work.

This is going to require a heavier use of tools, like NVIDIA’s Omniverse, to model out office space virtually to maximize the potential for future reconfiguration, so this office space can evolve over time along with the changing needs of users. It won’t be a specific configuration that will drive success, but the ability to rapidly change that configuration based on the changing needs of the business.

Flexible Spaces

As we move into increasingly using the metaverse and related tools to model both existing spaces and spaces in the design phase, we could design them to be highly flexible, so changing employee needs can be more easily reflected and addressed in the related architecture.

While virtual meeting spaces, like Meta’s Horizons, may not yet be up to meeting virtually as if you were in person, that technology will continue to advance. At some point, these meeting and social collaboration rooms may themselves become redundant. It’s possible that renting rather than buying new office space may be the safer short-term path.

The key analytics-based attribute for the new modern office is flexibility and fewer on-premises employee numbers than had been the case prior to the pandemic. Going forward, the most flexible offices will likely be the most successful, making building or provisioning for that flexibility the most important part of the effort.

IBM recently announced the massive Osprey, which is one of the most advanced quantum computers in the world. IBM also announced a partnership with Algorithmiq out of Finland that is developing a quantum simulation platform focused initially on health care and materials science.

The interesting result should be a significant improvement in related drug discovery efforts that, given quantum computing’s massive performance advantage with huge datasets, should help advance new drug development while significantly lowering side effects once finished.

The same problem that has plagued these efforts in the past, including access to data, particularly from research hospitals, hasn’t been fully mitigated. But federated and synthesized data efforts are slowly beginning to close those gaps to create the potential for that data to be available once a fully capable quantum computer can be spun up to the task.

Let’s talk about quantum computers and how they could significantly change the world and particularly health care this week:

The AI Edge in Health Care

The first time I was introduced to IBM’s Watson platform, it was focused almost exclusively on the medical industry. The M.D. who briefed me shared that once that old instance of Watson had been trained, he entered a series of symptoms from a woman he’d worked with for years to identify her illness. It took him around three years of focused research to identify a list of potential illnesses.

In short, even though this was a rudimentary form of Watson at the time, it changed a multi-year process into one that could arguably have been done in minutes. For many patients, it could cure an illness that might never have been diagnosed, given how much effort that diagnosis would have required.

Medical AIs require massive amounts of data to do their job, because they have to focus on the deep learning (DL) side of AI, given the high variability of both people and illnesses. Side effects, unintended adverse consequences, like addiction, and cost are all part of any effort to find an ideal medication to address a new or existing illness. Once mature and at sufficient size and scale, quantum computers will be able to deal with datasets that are far larger than we are able to realistically deal with, by using speeds that today’s conventional computers can’t touch.

This quantum capability should give IBM a significant edge in a market where these massive datasets and fast results are required and make IBM’s recent partnership with Algorithmiq critical to the successful future of the AI effort. In short, our ability to deal with a pandemic more effectively will likely be impacted by how mature this joint venture between the two companies is at that time. Once mature, it could be a medical game changer when it comes to developing better, safer, and more trustworthy medications.

Key Partnership in Health Care Market

IBM’s leadership in AI and quantum computing was highlighted both by the announcement of the powerful quantum computer and the announcement that Finland’s Algorithmiq would be partnering with IBM on drug discovery.

The combination of these two announcements showcases the very real near-term potential benefits for AI and quantum computing. Sometimes, having the right partner can lead to truly world-changing efforts. Finding a faster, better way to discover medications would go a long way to assuring longer lives and lowering our medical expenses over time.

As an OEM, Lenovo was all in early to embrace AMD’s performance advantages. Lenovo adopted AMD’s Threadripper platform for workstations and rode that decision to workstation leadership in the related segment.

At AMD’s event, Lenovo’s Kirk Skaugen shared that Epyc processors were number one in reliability and in high-performance servers, which speaks well of Lenovo’s aggressive AMD position on PCs and servers. Like the business PC unit, the server unit was acquired from IBM and showcases what can happen if two companies partner well with each other.

I had a chance to meet with Lenovo at the event, so let’s talk about AMD’s performance and sustainability advantages and what this will mean for Lenovo’s movement in the server market.

AMD’s Epyc Launch

This was a powerful AMD launch with many vendors coming on stage to sing its praises, such as Oracle, Super Micro, Dell, and Lenovo.

The performance and energy efficiency of the new Epyc line of server processors is competitively strong. AMD’s new-generational performance advantage is due partly to AMD’s unprecedented focus and execution by its executive team, particularly its IBM-trained top executives Lisa Su and Mark Papermaster.

I was at AMD’s gaming launch recently, and, as you would expect, there was a lot of rowdy behavior and cheering. At AMD’s server event, the crowd seemed as excited about the Epyc release as the other crowd did about gaming.

Lenovo AMD-Based Servers

Lenovo is an exceptional partner. It attends its partners’ events without pitching itself as being above the partner it’s supporting. Lenovo has been using AMD to aggressively move against its competitors successfully for the last several years.

Winning award after award and gaining in market share, Lenovo has also benefited from its Chinese roots, because to operate internationally, Lenovo was forced to implement extremely rigorous security controls, which have had the unanticipated effect of making it more secure and far more reliable.

Companies live on uptime, and based on survey results, Lenovo’s AMD-based servers are by far among the most reliable in the market. Lenovo also shared that the power savings alone from its latest AMD servers provide a 12-month full return on investment on energy cost alone. This is a testament to the pivot Lenovo made toward using AMD as a huge competitive differentiator. One of the most interesting Lenovo efforts is the line of Lenovo Neptune servers that use water cooling but don’t require an external water source. In this regard, they’re more like water-cooled gaming desktop systems and provide the benefits of water cooling without the extreme costs of plumbing the data center for water.

Today, Lenovo is aggressively leading the server market in a number of key metrics with strong growth internationally. Increasingly, what drives that growth is the synergy within Lenovo between units, suggesting we are still at the beginning of this growth spurt.

Lenovo-AMD Partnership

Lenovo is using AMD’s Epyc serve line technology and partnership to drive competitive advantage at a global rate in the server market. Sometimes you can take a risk, and it can make the difference between staying with the pack and market leadership.

Both AMD and Lenovo highlighted at the fourth-generation launch of Epyc that they are taking the latter path. The result puts both of their competitors on notice that they came to play, and they do not intend to lose.

For data centers looking to cut back on operating costs while improving performance, both AMD and Lenovo made a good case for getting your business. It’s worth checking out.

NVIDIA, which has been very successful with its own metaverse offering, went after business and government markets first, because those markets can justify the cost of what is currently a relatively expensive solution.

Microsoft’s HoloLens is a showcase for how, typically, a technology first rolls out where the money is. That way, the first instances focus on providing an adequate solution, which can then be cost reduced over time, then released to the consumer market when costs are low enough to justify a price that this market will accept.

In the consumer market, you still have the horse-and-cart problem of having enough content to make it worthwhile for that consumer offering, requiring the driving vendor to fund a lot of initial content, so people can see the benefit of having the related solution.

Dogfooding to market readiness

When NVIDIA first started selling Omniverse solutions, it was already using it internally. Omniverse is NVIDIA’s metaverse development platform.

The use of your own products is called “dogfooding” or “eating your own dog food.” The idea is that by using the offering yourself, you gain a deeper understanding of it and can better appreciate the problems and the priorities, knowledge that helps improve the product. But there’s one huge caveat: If the product does work well, you run the risk of crippling your workforce.

When working at one tech company, I first saw this with an operating system.. Few in the company wanted to use the product, and groups like marketing refused to use it, choosing Microsoft Windows instead. If your employees don’t want to use a product, there’s a serious issue. Ordering them to use what they feel is an inadequate product will only do more damage to the company.

The right path is to fix the product, so they want to use it, not force use. Because you can’t force a customer to use a product they don’t want. The employee is supposed to be a proxy for the customer, and they need to be treated as such, so the product evolves to be acceptable in the market.

NVIDIA’s employees, for instance, use NVIDIA’s Omniverse aggressively to design new buildings, train robots and cars, and better address global warming with the Earth 2 effort. The productivity benefits to employees were clear and the results very positive for NVIDIA which is the way these things should be done.

When it comes to dogfooding employees, they must want to use the offering. If they don’t, that’s the first problem that needs to be addressed, not by force, but by making the product more compelling.

See more: NVIDIA and How AI and the Metaverse Will Transform the Web

Executive vision

A CEO can’t use their power to force both internal use of an inadequate product and eventual adoption by consumers. The market doesn’t work that way.

NVIDIA’s CEO Jensen Huang, an experienced CEO, knows that if the employees don’t want to use something, it shouldn’t be brought to market until the product is altered, so employees want to use it.

Employees may get better support and lower prices, but if they don’t like the product even with those benefits, it is not market ready. Much like you can’t order a car to fly, ordering people to like a product they don’t like isn’t going to work either.

See more: The Metaverse is Overhyped; But by 2050, AI Will Make It Real

We not only have to balance often conflicting business schedules but conflicting life schedules. We have children, spouses, friends, and hobbies along with our work, and we all realize this is not necessarily conducive to creating the right work-life balance.

These conflicts force us to make hard choices that we often regret as we approach the end of our lives. This has certainly happened to me. We might not have been there to support a child, spouse, friend or coworker when they most needed us. Because we prioritized something tactically, like attending a critical meeting, and forgot that often it is our more strategic relationships that should have had priority. Thus, our relationships with family and friends have been weakened, particularly our relationships with our children.

There is a coming AI technology that, assuming we use it correctly, could make this problem obsolete. That technology is human AI clones or human digital twins. I am using the term human AI clone, because it better reflects what is coming and better focuses on both the promise and the problem with the technology. I will cover both the promise and the problem. But I would like to set the promise of the technology as the goal — and start having people think about how this technology should be used before we find it has been deployed badly.

Let’s begin.

The promise of AI employee clones or human digital twins

The idea of creating an AI-driven clone of a person is far from new. It is at the heart of digital immortality, and it will also be critical for the creation of large-scale environmental simulations, be they focused on industry or entertainment. On the latter, the concept of realistic non-player characters (NPCs) would be a game-changer. But the true benefit to us individually is that we can delegate an increasing portion of our lives to the clone.

How much do you contribute to a meeting? That depends on the meeting. In large meetings designed to provide updates, we are often merely observers and provide no additional value. In other meetings, we may ask one or two questions, and on way too many occasions, we find we did not even need to be there. However, we seldom know this critical piece of information until we are in the meeting and see the content. Virtually no meeting leaders send their content out before the meeting. Often, even a full agenda is the exception rather than the rule.

One fix is to demand better meeting preparation. But we’ve been pushing for better meeting preparation for as long as I have been in business, and we just have not made that much progress. Granted, it could be because too many people wait until the last minute to work on their content. But if we can’t fix this problem one way, we can use technology to fix it another way. That problem is that our time is a limited resource, and we need better ways to focus that limited resource where it can be most effectively used.

This is where the AI employee clone could provide the best benefit. It could attend the meetings, respond to generic queries as our proxy. And if it is hit with something it hasn’t been trained on or it needs other help, it can contact you real-time with a query, and you could advise how it should respond on your behalf. You could even step in quickly to provide direction. There are still reasons you should attend in person, but this would provide you with the choice of whether you need to attend, while today, you generally don’t have that choice. Assuming you use that choice wisely, this could be a path to both better work-life balance and higher-focused productivity. We could spend more of our time where we are needed and waste far less of that very limited resource.

See more: The Top Artificial Intelligence Companies

Preventing replacement

The problem is where I fear this technology will be focused, which is employee replacement. That would be a bad outcome: because it would cause employees to fail to train their clones; and because it would feed into too many executives’ belief that companies would run better if they could just eliminate their employees.

These human AI clones need to be owned by the employees that create them, and if they are used after the employee leaves the company, the employee should continue to be compensated. This could flow into their eventual retirement, creating a stronger reason to invest the time into creating that tool.

Assuring this technology benefits rather than harms the employee is critical to assure that it will both do what it will be able to do and drive the huge potential work-life and productivity benefits it can provide.

See more: AI, Automation, & the Future Impact on Jobs

What side is the money on?

The promise of employee clones is significant. It could mean more time spent on things we love to do at work and far less on things that waste our time and annoy us while improving work-life balance and helping to fund our retirement.

The problem will be the desire for companies to own the resulting clones and use them to replace their employees. Sadly, the money is on the wrong side of this, and our goal is to assure that this money dynamic does not drive the wrong outcome.

See more: Artificial Intelligence Trends & Predictions

The ancient military strategist Sun Tzu argued “to win a hundred victories in one hundred battles is not the acme of skill. To subdue the enemy without fighting is the acme of skill.” So in war, the way to avoid what is strong is to strike what is weak.

When comparing how Cyndx is competing in the search market and how Zoom is competing in the video conference market showcases one company going after where the market is weak and the other attacking where the market is strong.

Cyndx, a focused search engine supplier for investment banking, is approaching the fight like Tzu would by hitting the market where there’s an opening with a targeted service, while Zoom is developing a competitor in one of the market’s strongest product categories.

Let’s explore how Cyndx is approaching what should be an impossible task successfully with an aggressive use of AI, while Zoom is taking a different approach:

Cyndx

What Cyndx has done is create the kind of tool that search market should have created but didn’t. So Cyndx was able to successfully be a dominant search tool in many banks.

Banks invest in a lot of technology but often don’t understand that technology well enough to optimize it. That means that related implementations in banking often perform well below their potential, and with investment banking, where you need as much information as possible about a potential investment, this results in avoidable process failures and losses.

Using advanced artificial intelligence (AI) techniques, Cyndx provides a focused tool specifically designed to rapidly execute diligence and aggregate and curate the information surrounding a potential investment, so the financial analyst doing the work gets a quick, thorough assessment of all related data and can then make an informed decision about the acquisition. This provides multiple advantages both in terms of the quantity and quality of the information supplied to the decision maker, so they can both move more quickly and with far greater assurance than they could using standard search.

Covering seven languages, the embedded AI in the Cyndx search tool uses what it knows about prior users’ similar searches to help refine and focus the financially related search and incorporate elements, like company size, number of employees, and location, both into the search criteria and presented as results depending on need.

Typical results include major suppliers, significant customers, curated patent data — which showcases where the company actually spends its time, and the validity of those patents — are they defensible, have they been successfully defended — and executive staff backgrounds. A recently added feature is an instantly derived valuation, so the investing entity can determine if an investment is supported by the target firm’s valuation.

This tool is actively used in around 90% of investment banks but only in around 10% of the companies actively doing acquisitions, suggesting both a growth opportunity for Cyndx and a competitive opportunity for non-banking firms actively doing similar types of investments and acquisitions.

This is a subscription service, so it isn’t corrupted, and it focuses on an area where the search market isn’t, which means the market likely isn’t, and won’t, work on coming up with a competing offering.

See more: Artificial Intelligence (AI) in Banking

Zoom

Zoom is rightly concerned about competitors bundling videos conferencing with productivity suites. However, collaboration is an obvious feature to add to a productivity suite, making it hard and likely inadvisable to try to use government ant-trust tools to fight.

To address this, Zoom has signaled they are expanding into the productivity space, mirroring what some tech companies did when faced with similar issues.

Zoom’s approach, therefore, could be better served by using Cyndx’s approach. Rather than attacking the productivity market, going after specific market segments, like education or remote meetings for governments or medical collaboration — requiring high security but with little need to integrate with office tools — and tech platforms — like Linux, Android, and Chrome — are rare moves that Zoom could defend. They may fly under the radar and allow Zoom to be more successful.

Regardless of these moves, Zoom should be advancing its AI capabilities to better mirror what competitors are rolling out to reduce the frequency of competitive migrations.

See more: Zoom Product Overview

Two market approaches 

When going after a larger competitor, remember Sun Tzu’s advice. Do a competitive assessment to better understand not only where the competitor is weakest but where it currently isn’t focused.

Cyndx has clearly done that in the search market. Cyndx is not only expanding where there is market opportunity, it is likely the market doesn’t care, meaning it won’t resource displacing Cyndx.

On the other hand, Zoom appears to be attacking one of the areas where the market is strongest and doing so in a way that will focus the market on efforts to move aggressively in return, resulting in an expensive competition with companies with great resources and market presence.

In the end, the lesson is that Tzu was right: If you want to win, understand your competitor and go where it isn’t and try to specifically find approaches that won’t focus them on your market to reduce your competitive risks and cost.

See more: Why NVIDIA Has Become a Leader in the AI Market

While I got particularly excited about Microsoft Designer with DALL-E 2, there were a ton of other AI efforts — that surrounded everything from development to video calls — that will change the way we work and collaborate.

Let’s cover some of the highlights:

Two paths for AI

There are two potential paths for AI.

One is to replace humans and use AIs instead of people. This path is consistent with the robot that Elon Musk announced that is designed to physically replace a person.

The other path, which is almost globally supported by tech companies like IBM and Microsoft, is to create AIs that assist employees and make them more productive and capable.

It isn’t hard for me to support the latter path, because, like you, I’m not ready to be replaced by an AI robot. But I often struggle with the fact that a lot of the work I do is repetitive and could be more easily done by an AI, leaving me to focus on what I enjoy doing.

Filling out forms, creating surveys, finding pictures I don’t have to pay for, and having a video conferencing session that feels more like it was produced by pros where I look great — as opposed to the way it is now where I can look distracted and not engaged even though I am — are just a few of the things I look forward to passing on to AIs.

See more: The Top Artificial Intelligence Companies

Low code to no code

I was so excited about learning how to code until I had to do it. I knew what I wanted to do, but the effort to create a viable application was highly repetitive and detail oriented. I liked the conceptual phase of the process, but the execution phase, not so much.

At Ignite, Microsoft showcased how it’s driving AI into tools like Git-Hub Co-Pilot and Power Apps, by lowering the amount of actual coding a developer needs to do and eventually, putting the creation of the application within the control of the user.

Now, I want to share my first large-scale coding effort as a user. I was in the finance part of my executive rotation. We had to learn every major aspect of the corporation, and I needed a CRM application. This was before the term “CRM” was created. I sat down with IT and explained what I wanted. Months later, they delivered a solution that was harder and more painful to use than using the desktop apps I’d used previously. It was put in place by policy, but the issue was that the developers I worked with had no real concept of the job I was trying to do and didn’t appear to care.

Over the years, I’ve seen this same problem recur almost every time an operational unit must get an internal or external developer to create a new application. The lack of fundamental understanding for the job or task tends to result in an initial offering that is unacceptable, followed by a long number of revisions that improve the app over time to minimally meet the initial need.

But how we better dealt with this problem back in the day was to use Office or similar tools to create the app ourselves. While the result wasn’t as refined as what came out of IT, it tended to work on day one and tended to improve to the limit of the tool, most often Excel.

See more: The Top Low-Code Platforms

Rogue apps

The capability to create working apps just by describing what you want, once it is easy enough for users, will create a subsequent problem, which is to make sure the resulting apps comply with policy, particularly the policies that correspond to security. These apps will be widely shared but have the potential to do massive damage if they aren’t fully third-party tested and secure. I expect a lot of apps will circulate outside of the control structure, creating avoidable operational and security problems at an impressive rate.

Assuring expedited and easy approval processes are in place and affirmed by policy will be critical to making sure this amazing new capability doesn’t become an even more amazing large-scale problem. Doing one without the other will be problematic. App review and approval tools and processes will need to advance but not become significant impediments. Otherwise, users will bypass them, creating a nightmare of unapproved deployed apps that are both against policy and too beneficial to the company to easily shut down.

Microsoft’s AI future

From creating pictures and apps from words to answering difficult operational and HP problems just by asking an appropriate question, Microsoft is imagining a future significantly enhanced with AI capabilities.

But this pivot will require users to develop descriptive skills. Much like those of us that use the web needed to learn Boolean logic, users of these AIs will need to learn how to most effectively interact with them. Being descriptive and clear in our communications is something taught to a minority of coders and students in general but that’ll need to be fixed if this effort is to be successful. Because if you can’t accurately describe what  you want, you won’t get it.

What will help is that AIs increasingly will learn how to translate your vocalized intent ever more accurately. But if you can do it better than most at the start, you’ll shorten your time to productivity and increase  your value to the company significantly. Me, I’m looking forward to creating my first picture with DALL-E 2 in Microsoft Designer.

But we are about to be up to our armpits in AIs and especially user-created code, so putting in place policies to deal with the latter will likely serve you as we approach our AI-driven future.

See more: Microsoft Azure Artificial Intelligence (AI) Portfolio Review

Intel has one advantage in the market that Koduri is going to be betting on: It doesn’t have any high-end cards to protect. 

This means that the typical strategy of putting the most compelling features into your high-end offerings and leaving them off more value-priced efforts to protect those higher-revenue and -profit cards isn’t a strategy that Intel will execute. 

Thus, compelling capabilities that only exist in those high-end cards, strong upscaling and artificial intelligence (AI), will be available in Intel’s more attractively priced alternatives. The competitors can’t respond to this move with a like move, because they still have those high-end card margins and prices to protect. 

This turns the typical tiered pricing and product strategy on its head and allows Intel to argue you are getting much of what you’d want in an higher-priced card in a lower-priced card.  

Intel’s marketing risk

An issue for Intel will be funding the needed demand-generation marketing program that will clue users in on this value, because funding a marketing program for a value-priced part is problematic. 

If users figure this value out, it could result in a shift in market share to Intel. But that also means many users will have access to these upscaling and AI capabilities that they would not have otherwise been able to afford. 

This means that, if successful, Intel’s move into the GPU space will enable far more hardware-based AI capability at the edge. That capability will drive changes in art, science, and education.  

PC AI and productivity

Applied AI will be able to provide much better editing and creation tools. 

Intel demonstrated its text-to-art capability on stage. Using its GAUDI2 platform, Intel was able to get the AI to create a rich and accurate image by describing what they wanted the computer to create. 

Imagine being able to create your own images for papers or books, without having to worry about running down the artists or being legal issues related to images. Better, because most existing images will only approach your vision for the piece, GAUDI2 can create the perfect image to accompany your work.

AIs aren’t just useful for creating images. They are already being used to write papers from brief summaries, dramatically reducing the time it takes to issue reports. Coupled with the image creation capability, it isn’t hard to imagine a future when the need to write most papers will be passed on to our computers, which, increasingly, do both a better and faster job.  

Finally, this potentially enables the market we’ve been waiting for: intelligent digital assistants, a concept that has been presented by IBM but has not been delivered in any kind of scalable project.

This movement of GPU-based AI capability into the low end of the mid-market should provide more people with the potential to have a truly capable AI that can carry out tasks and provide answers that are more comprehensive and accurate — and even increasingly able to carry out scheduling and traveling efforts automatically.   

Disrupting the PC market

Intel’s move with its ARC graphics platform will focus compelling features, like upscaling and hardware-based AI, into more affordable PC and laptop lines. 

This will increase the potential to sell apps based on these technologies to more people at a far more affordable price. It will also make the market for AI-based apps more attractive because of the related higher sales volumes. Intel is executing a good strategy against entrenched vendors that have to protect their high-end offerings. 

Intel is releasing ARC to desktop and laptop PCs — including ARC-based laptops by Lenovo and Samsung — which will help the AI PC revolution begin once those sales get to critical mass. 

Oh, and it is also being focused both on creating malware that would be very hard to stop and tools to stop that malware once it invades your space and on ever more capable autonomous weapons systems.

But it strikes me that our initial priorities should be focused more on correcting endemic problems rather than effectively creating new ones, like massive unemployment, AI-driven fraud and malware, and weapons.  

Let’s look at three missing strategic opportunities for AI:

1. Helping kids make better decisions

When most of us were kids, we went to school because we had to. We had no deeper goal nor did we have any idea what kind of career would benefit us. 

Many of us came out of college with near worthless degrees in subject areas like literature that had little relation to what we wanted or needed to have to create a career we’d enjoy. In fact, I’d argue that most of us start our careers with little or no idea of what we like and dislike in a job. Instead, we learn that over time as we advance out of our first jobs in college.

And the lack of STEM students is severely hampering our ability to compete on a global scale, with way too many people coming out of colleges and universities with skills the market doesn’t need and too few with skills it does need. 

One obvious use of personal AIs would be to use what has been learned by those that enjoy their careers to connect the dots to kids that have similar behavioral attributes and interests. This could enable them to fast track into a job and career they are more likely to enjoy. Their education would be focused not on merely completing that education, but on gaining the critical skills they’ll need to be successful in a career.  

2. Bad idea protection

We are surrounded by digital assistants, some of which observe and most of which listen to us during the day. 

We often make mistakes, because our actions exceed our ability to think through those actions timely. Road rage, an inappropriate post on social media, or an inappropriate joke all can cause catastrophic and long-lasting consequences that we should want to avoid.

A monitoring AI that was trained on past bad behaviors could alert you that you are in the process of making a huge mistake and even actively prevent you from making it by disconnecting your keyboard, smartphone, or disabling your car to protect you from yourself.  

Many of us can be our own worst enemy and having an AI that has your back could go a long way to assuring you don’t accidentally destroy your career or life. 

3. Relationship guidance

Those of us who have happy relationships seem to be in the minority, and it strikes me that there are a lot of people that should never get married for their own good. 

Again, an AI that has been trained to identify bad relationships and on identifying behaviors that destroy relationships could go a long way toward helping people find better mates and, once found, make sure the relationship lasts. Simply sharing the downside of a work relationship might help prevent one that would otherwise turn out to be career or life limiting. And it could alert those who shouldn’t get married that doing so will most likely result in a ton of regret. 

AIs should be focused on helping ensure that not only are both parties compatible, but both are on the same page when it comes to managing a relationship, so the typical expensive breakup or divorce, rather than being almost certain, becomes exceedingly rare.  

Using AI to help us

Artificial intelligence is one of the major new tools that will define our future. Yet, it isn’t focused enough on assuring we want to live in the future that it defines. 

There is a significant risk that future AIs will be hostile to humanity and the best way to avoid that outcome is to over pivot on assuring they are designed to help people, rather than to replace or sell to them. Helping people make better life decisions can go a long way to both improving their productivity and happiness, benefiting us far more broadly than AI robots that replace us, sell to us, or are designed for weapons.  

If we can instead focus the AI on helping kids make better career-oriented decisions, protect us from our own stupid ideas, and assure both our relationship skills are stronger and longer lasting, we could go a long way towards assuring a brighter future for all of us. Sadly, I doubt we will see this priority shift any time soon.