SoundHound AI has enhanced its AI-based food ordering capabilities with the acquisition of Allset, an online ordering platform that connects restaurants and local customers.

Financial terms of the deal, announced Thursday, were undisclosed.

Allset is a food ordering platform designed for local pick-up, working with nearly 7,000 restaurant partners nationwide, including Joe & The Juice and Charleys Cheesesteaks.

Speech-to-meaning

SoundHound, a voice AI and speech recognition company, was founded in 2005 and went public in 2022. The company develops speech recognition, natural language understanding, music recognition, and search technologies.

In 2023, SoundHound generated $45.9 million in revenue, a 47% year-on-year increase. The vendor is growing but not yet profitable.

As part of Nvidia’s broader strategy of investing in AI-related tech companies, the chip maker has invested $3.7 million into SoundHound, in return for a 0.6% stake.

SoundHound’s Vehicle Intelligence product, powered by Nvidia Drive, allows drivers to ask questions related to maintenance, safety, and other vehicle-specific information using natural speech.

SoundHound markets itself as an independent voice AI platform, offering customers the ability to create their own branded voice experiences rather than relying on voice assistants from bigger vendors.

Allset

SoundHound said the deal to buy Allset will advance its plans to enable voice-enabled food and drink ordering across millions of cars, TVs, and smart devices. Its voice ordering technology works across multiple channels, including via phone, drive-thru, kiosk, and mobile app.

More than 10,000 restaurant locations use SoundHound’s platform to understand speech in a range of major languages, learn any restaurant’s menu, process orders directly to the point of sale (POS), and answer customer FAQs.

“Allset will help SoundHound bring voice AI solutions to even more restaurants looking to improve operational efficiency,” Keyvan Mohajer, CEO and co-founder of SoundHound AI, said in a statement on the deal.

For example, SoundHound has been working with White Castle, the US-based fast-food hamburger joint, to offer voice AI ordering technology at select White Castle drive-thrus for a year, with plans to roll out the technology to 100 locations by the end of 2024.

Many in the fast-food industry are looking to use AI technology to increase efficiency and reduce costs.

Using an AI system that provides a consistent, interactive ordering experience is a big ask.

Super size snags at McD’s

McDonald’s recently ended its trial of AI-powered drive-thru technology developed in partnership with IBM. McDonald’s had been testing IBM’s AI-powered voice recognition technology at around 100 US drive-thru locations for around two years.

The technology has numerous glitches, chiefly around the misinterpretation of customers’ orders to sometimes comical effect.

Hard to swallow mistakes included adding bacon to ice cream and putting excessive quantities of items on orders (e.g., hundreds of chicken nuggets on a single order). People took to social media with tales of “fighting the McDonald’s robot” and similar.

The automated order taking system — which evidently had problems picking up on accents and dialects, noise and cross-talk from neighboring drive-thru stations — will be dropped from the end of July, Restaurant Business reported.

IBM has yet to respond to a request for comment from Computerworld but reportedly said it looks forward to working with the fast food restaurant chain on a variety of projects in the future.

Bill Conner, CEO of mobile technology developer Jitterbug, told Computerworld that teething problems with new technologies are to be expected. Waiting for technologies to mature is unwise, since slow adoption of AI technology is likely to leave businesses at a competitive disadvantage, Conner argued.

“The future of application development, orchestration, and automation is based on an AI evolution, not a revolution,” Conner said. “Even the most agile organizations need a smart, measured approach to infusing AI capabilities into their business and infrastructure.”

As employee experiences become increasingly digitized, the digital employee experience (DEX) is becoming the primary interaction you and your workers have, whether in person, remote, or hybrid. That’s why the technologies that enable whatever mission you happen to be on have become so important, as the tech is by definition an essential component in any digital employee experience.

We know because they told us

We already know this because big enterprises like SAP, Salesforce, Cisco, IBM, and many others have told us that when given the choice, employees will choose an Apple product as their primary workplace device. We also know that, for example, sales of portable Macs absolutely boomed during the pandemic, when every office-based worker transitioned to become a home-based employee in a matter of weeks. 

Once the pandemic seemingly ended, those home-based workers became mobile employees, and later — sometimes reluctantly — turned into hybrid workers. 

In that sense, the importance of Apple’s products has already been proven. These devices, whether Mac, iPhone, or iPad, have already been used on a global scale to maintain businesses remotely in real time — as will visionOS devices in their time. 

That success continues to translate into increasingly large deployments across industries you might not have considered to be natural Apple users before, from retail to engineering and beyond. 

Empowering good business with DEX

Of course, business leaders recognize the changing workplace, with 24% of leaders already seeking to unlock improved DEX and productivity. The analysts at Gartner say digital workers who are happy with the tools and applications provided for work are 1.6 times more likely to stay.

While not everyone is an Apple user, sundry TCO studies suggest that businesses that do adopt those products realize significant benefits in terms of support costs, device longevity, employee retention, and productivity.

But at its core, Apple’s key offer to business is the same one it offers to consumer users: a user interface that, for the most part, gets out of the way to enable the user to get what they want to get done as effectively as they can.

That’s the whole point of smart DEX strategy — to equip and empower employees so they can stay focused on their task and not get bogged down by tech.

This secret sauce is compelling to any busy person, and the savor gains gusto on strength of Apple’s full platform — by which I mean that on mobile, tablet, or computer you get a similar DEX.

This familiarity is critical, as it makes it easier to complete tasks on whatever device makes the most contextual sense for a person at any point in time. In essence, the provision of consumer-simple technologies capable of delivering enterprise-class results to employees who can work from anywhere empowers immense business flexibility. 

Infrastructural change

The other post-pandemic benefit supporting the trend comes through the rapid evolution of device management of Apple devices since the pandemic first hit.

Sure, there was already a healthy industry of device management vendors (including Apple itself), but the company took note of enterprise pain points and introduced tools to address them. Think zero touch, declarative device management, improvements to activation lock, and even the recent introduction of a dedicated (and provisionable) password app based on the company’s existing iCloud Keychain. 

The result? 

At the front end of business, DEX improves through employee-driven adoption of Apple’s products.

But the company has also been speaking with both Windows- and Apple-familiar IT admins to figure out how to enable good experiences for them. It’s now truly possible to ship hardware to new hires, who can safely and effectively start their managed employee experience with a single login to their new device — and for IT to manage the entire transition through a single pane of glass.

Please follow me on Mastodon, or join me in the AppleHolic’s bar & grill and Apple Discussions groups on MeWe.

One small but key new feature for iCloud in macOS Sequoia should make developers and engineers a little happier, as it will be possible to sign into iCloud and other Apple ID (soon Apple Account) services from virtual machines.

This hasn’t been possible before, which has impeded developers and users hoping to test iCloud features in apps, including trialing iCloud sync on their software. It has also been an obstacle for any Mac user who just wanted easy access to iCloud data form within a virtual machine. 

Who is this feature for?

For most developers, the big advantage will be in software testing. Many developers need to test software in multiple Mac configurations, but not all of them have the fleets of Macs required. Given that so many app developers now make use of iCloud features or make use to some degree of iCloud sync (think calendars, for example), the inability to easily make such tests may impact customer experiences.

Developers who like to keep a firewall between clients and their projects by hosting these in different virtual machines should also benefit from this support, as it means their iCloud data and services remain consistently available independently of their workflow.

So, what is the deal? 

Apple says that so long as both the host machine and VM are running macOS 15 Sequoia or later, it will be possible to use iCloud and other Apple ID-related services with the hardware. It’s not a complete panacea, unfortunately. That’s because the need to run macOS 15 or later means developers wanting to trial older operating systems will still have problems. All the same, it means that over time, it will become possible to test the performance of older macOS versions as new iterations ship.

There are a few technical considerations, as explained in an Apple tech note:

• If someone moves a VM to a different Mac host and restarts it, the Virtualization framework automatically creates a new identity for the VM using the information from the Secure Enclave of the new Mac host. The person using the VM will need to reauthenticate to allow iCloud to work.
• The Mac will also detect attempts to start multiple copies of the same VM simultaneously on the same Mac host. 

Enterprise-ready Macs

This is just one of a slew of WWDC-announced improvements coming to Macs and other Apple devices this fall. These include improvements in device management, passkeys, software updates, and Safari management.

Admins can also install executables, scripts, and configuration files remotely via MDM, which should make it a little easier for organizations to deploy and manage services across their fleets. Apple Account (formerly Apple ID) management has also been improved to encourage organizations to use Managed Apple Accounts..

Another great improvement is in Activation Lock. There have been problems in which perfectly functional Macs that have been handed on for renewal or sold legitimately on second-hand markets have ceased to function because people have neglected to disable Activation Lock. Apple does have a process to help in cases of this kind, but at WWDC it introduced new support to enable IT admins to remove Activation Lock using Apple Business Manager, which should mitigate the unintentionally bricked Mac problem.

Finally, Declarative Management will replace MDM profiles for software updates, as well as for some other commands and queries. The beauty of this is that Declarative Management delivers far more transparency to admins, while also improving the experience for users — your Mac won’t suddenly enter an update cycle just as you sit down with the board to begin a critical business condition status presentation, for example.

One more thing: whispers tell me it is also going to be possible for IT to disable services such as access to OpenAI from within Apple Intelligence, so unapproved sharing of enterprise data should not happen on managed devices.

All told, WWDC has been really good for IT admins, who now enjoy access to a broad and growing range of powerful tools with which to manage their Apple fleets. This seems particularly timely as the size of those fleets continue to expand. 

Please follow me on Mastodon, or join me in the AppleHolic’s bar & grill and Apple Discussions groups on MeWe.

Slack CEO Denise Dresser this week laid out what she sees as her company’s primary use for artificial intelligence (AI) in the future: making it easier for users to find key channel conversations and turn them into the basis for tasks and projects.

Dresser, a longtime Salesforce employee who took the reins of Slack six months ago, sat down with members of the media on Tuesday after her keynote address at Salesforce’s World Tour event in Boston. Much of the focus was on Slack’s integration with Salesforce’s Einstein Copilot.

(Salesforce acquired Slack in late 2020.)

What Slack will eventually be able to offer both its own and Salesforce’s users is a unified experience where AI oversees any influx of both structured and unstructured data and parses through it to offer users the most important summaries. Being able to find key moments in chats and knowing what happened in conversations is hard to navigate, Dresser said, and is at the heart of Slack’s AI integration.

“Sometimes AI can be the simplest thing that drives productivity,” she said. “So, we did Slack AI Search. With that, Slack becomes the long-term memory of your organization. …Being able to find things easily in a generative manner, where you actually get a summary of what you need to find, was a really big ‘Aha’ moment for us.”

AI’s adoption and integration into virtually every Slack function will continue to accelerate. Dresser pointed to an “evolution of skills” that has come with the adoption of the technology, including prompt engineering or the use of natural language processing to perform functions, such as the creation of software without traditional line-by-line coding.

“It took two months for ChatGPT to get to 100 million users, 15 years for the mobile phone to get to 100 million users, four years for Facebook to get to 100 million users. What I think we’re going to start to see is this [same] acceleration as people start to adopt it, and see productivity improvements,” Dresser said.

“We’re going to bring that into Canvas. It’s going to be in Workflow, it’s going to be in Huddle,” she continued. “So, you’ll see AI infused everywhere. It’s just going to be by your side in the application.”

Users, Dresser explained, won’t even know it’s AI with which they’re interacting; it will be a natural offshoot of the Slack functionality. For example, users would need to use a search window to weed through days of Slack messages they may have missed. Instead, an AI-infused Slack would quickly surface the most important message summaries.

In terms of future innovations, Dresser pointed to the recently launched Slack Lists feature, which automatically captures the most important parts of channel conversations and surfaces them to users.

Less than 34% of projects are completed on time and one budget, Dresser claimed, saying that users having to switch between tasks in applications was a significant drain on time and productivity. “We have millions of people working in Slack; why leave Slack?” she said. “We wanted to bring that capability for tasks, and lists, and projects into Slack. It starts right in a [Slack] conversation, where you’re able to start a task list from that conversation and start working on your project right there.”

The AI-infused communication and collaboration platform will eventually also suggest to its users the chat channels they should prioritize for project purposes. “That type of power in terms of capability is going to be ‘Aha’ moments for people,” Dresser said.

She noted that only about a third of employees in general use AI-powered platforms in their jobs — but those who do have seen an average 81% productivity increase by eliminating mundane tasks.

As AI continues to be integrated into Slack and Salesforce tools, one challenge will be maintaining the feel and “integrity” of what’s she called a “beloved” application.

“We’ve already integrated Slack, Sales Elevate, and Salesforce. Copilot’s integration is going to be great,” she said. “One of the things we thought deeply about was making sure the craft of Slack and the experience of Slack is maintained, even when thinking about architectural integration. Creating that experience that is very Slack-like and that’s efficient and productivity is something we’ve thought deeply about.”

The US is once again trying to slow advances in China’s chip manufacturing industry, a move that is unlikely to pay dividends long term, but may help US business concerns in the very short term, according to analysts and other China technology experts. The outcome will shape the semiconductor industry in the years to come.

Reuters reported on the latest diplomatic efforts on Wednesday, noting that US export policy chief Alan Estevez is traveling in an attempt to expand a 2023 agreement between the US, Japan and the Netherlands “to keep chipmaking equipment from China that could help to modernize its military.”

But experts on Chinese technology strategies are skeptical the talks will yield much and, if anything is achieved, they are even more skeptical that it will help long-term. That said, even a short-term delay in China’s chip-making progress could give US companies more breathing room.

The negotiations are tricky. US negotiators must convince key governmental and industrial concerns to stop selling chip manufacturing equipment and materials to China, even though it means surrendering a massive amount of revenue.

Lithography systems a key lever

Japan’s chip strength comes from a wide range of materials and components, whereas the Netherlands’ specialty is lithography systems, the complex machines used to etch semiconductor designs on silicon wafers, said Mario Morales, the group VP overseeing semiconductor coverage for IDC. Lithography “is what makes the most advanced silicon today. These are all critical market segments, so it definitely slows things down, but China is absolutely investing in those spaces,” he said.

The sums of money at stake make the US diplomatic efforts unlikely to succeed, Morales said.

“Some 25 percent of Intel’s business is coming from China,” he said, adding that AMD, Nvidia, and Qualcomm also rely heavily on sales to China. And if the US negotiators are successful, Morales said that he expects China to retaliate.

The only argument that the US can make is long-term, Morales said. These companies all know that China is heavily investing in its own manufacturing operations, and it will only be a few years — Morales projects five to seven years — before China will no longer need materials from Japan and the Netherlands. Once that happens, China will cut off revenue to those companies anyway.

The argument goes that not cooperating with the US now may make the US less likely to be there for those companies when China no longer needs them. On the flip side, it’s unlikely the US would cut back on purchases from Japan and the Netherlands even if they don’t cooperate because US companies need their support. It’s not as though the US would want American companies to buy from China any more than is absolutely necessary.

Morales specified automotive, smartphone and China’s own AI operations as representing the bulk of China’s current chip investments. “These are all critical market segments, so it definitely slows things down, but China is absolutely investing in those spaces,” Morales said. “Most of these initiatives will fail. And even if they don’t, it only contains China for a short term.”

Irina Tsukerman is a geopolitical analyst, a national security lawyer and the president of Scarab Rising, a global strategy advisory firm. Tsukerman argues that these talks are going to make things very awkward for some of these chip component and tooling manufacturers.

Short-term profit or long-term relationship

“The scramble over these restrictions means companies will hard pressed to make a choice whether to take maximum advantage of that window of opportunity to make profit, likely at escalated prices, to sell that equipment to China,” Tsukerman said, “or get on the US good side early and voluntarily start cooperating with the additional restrictions with a possibility of gaining early leverage and scoring investments or other positive incentives from Washington, before negative incentives are introduced more broadly.”

But she agrees with the US efforts, as she argues that it will likely succeed in the very short-term.

“Until now, China has been able to capitalize on weak import-export controls, including via US-friendly countries, as well as loopholes in the type of equipment banned under restrictions, to continue apace with its development of its domestic chip industry,” Tsukerman said. “Netherlands and Japan remain in possession of some of the most advanced equipment for developing semiconductors. If these loopholes and restrictions are tightened, China will have a much harder time with making the breakthroughs it needs to circumvent the overall bans on AI chips.”

Tsukerman said that US government officials should also look inward, as China has taken advantage of ineffective enforcement from multiple US agencies. The US “should look towards its own import-export controls, because the occasional and sometimes systematic unwillingness of its agencies to enforce these controls has resulted in China exploiting additional loopholes to gain access to such equipment,” she said. 

Retaliatory measures

In terms of likely Chinese retaliation, the Chinese government would have many options. “The US and its allies should be preparing for a likely eventuality of China placing additional restrictions on the rare earth minerals and other raw materials used in the development of semiconductors to prevent the US and its counterparts from advancing further and making it easier for China to catch up even under restriction,” Tsukerman said.

Potentially more importantly, Tsukerman said the US focus on China might prove problematic, as these chip supplies could also come from various other countries. “Working only with the Netherlands and Japan to prevent China from gaining access to equipment is clearly insufficient. China has benefited from loose import-export controls in countries willing to sell that type of equipment produced with Dutch, Japanese, and/or joint US involvement to China with the same result. The coalition should work in tandem to shut down gray zones, possibly imposing secondary sanctions on countries willing to sell such equipment,” she said. “For now, the chipmaking equipment production is dominated by the US-led coalition. It is only a matter of time before other countries develop their own capabilities, BRICS members being the prime example,” she added, referring to the intergovernmental organization founded by Brazil, Russia, India, China and South Africa and more recently expanded to include Egypt, Ethiopia, Iran, and the United Arab Emirates.

Tsukerman said recent US efforts to negotiate with India may be a good move, “India could still be helpful to chipmaking production in any of the expanded BRICS members in many other ways. Meanwhile, such restrictions certainly do not apply to any of the other members, including Saudi Arabia, which is dedicating billions to start a regional semiconductor hub and may receive China’s assistance in exchange for providing China with such equipment.”

Brian Levine, a managing partner with Ernst & Young who was one of the US Department of Justice’s representatives in the US law enforcement Joint Liaison Group (JLG) with China, said that he sees even a miniscule delay in China’s chip-making efforts as worthwhile for US interests. 

War is waged with microchips

“Progress is progress and the same is true for the delay of progress. These days, war is waged not with guns, but with microchips,” Levine said. “Win the war on microchips and you may win the war generally. I think the administration is trying to pull what levers it can. The degree to which it will be effective will depend on many factors that are unknowable at this point, such as the level of international cooperation with the effort.”  

Although some have raised questions of hypocrisy with the US trying to pressure China into not doing what it is actively doing, Levine disagrees. “Much of this technology was likely stolen from the US and other countries, so I have limited sympathy for those who may get worked up about the US attempting to impose such restrictions.”

Robert Khachatryan, the CEO of Freight Right Global Logistics, is also uncertain about how much of an impact the US diplomatic efforts will make. 

“Although the restrictions may slow China’s progress in the short term, it is uncertain if they will be sufficient to halt it entirely. China has been increasing its investments in domestic chip production, aiming for self-sufficiency,” Khachatryan said. “Restricting China’s access to AI chip technology could disrupt global supply chains and commerce, as China accounts for 24 percent of global semiconductor consumption. Such restrictions might slow worldwide production and distribution and shift trade dynamics, with countries seeking alternative suppliers.”

Microsoft has added generative artificial intelligence and other enhanced features to its quantum-computing platform as part of a larger strategy to deliver the game-changing technology to a broader range of users — in this case, the scientific community.

The company on Wednesday unveiled the release of Generative Chemistry and Accelerated DFT, which together expand how scientists in the chemicals and materials science industry can use its Azure Quantum Elements platform to help drastically shorten the time it takes them to do research, the company said in a blog post.

“Just as generative AI has unleashed new waves of creativity and improved productivity with collaborative tools like Copilot, we are now bringing AI and natural language processing capabilities to science,” according to the post, attributed to Jason Zander, EVP, Strategic Missions and Technologies.

Shipments of augmented reality and virtual reality (AR/VR) headsets dropped 67.4% year over year in the first quarter of 2024 as a result of an evolution in the market, new data from International Data Corp. (IDC) reveals.

“The decline in shipments was expected as the market transitions to include new categories such as Mixed Reality (MR) and Extended Reality (ER),” IDC noted Tuesday. “Despite the decline, the average selling price (ASP) rose to over $1,000 as Apple entered the market and incumbents such as Meta focused on premium headsets such as the Quest 3.”

The future of such products in the enterprise is in flux, with Microsoft pulling back and laying off workers from its HoloLens division last year, while Apple is clearly targeting the enterprise market with its Apple Vision Pro.

The research firm said that it recently revised its taxonomy of headsets to incorporate two new categories: “Mixed Reality which occludes the user’s vision but provides a view of the real world with outward facing cameras, and Extended Reality, which employs a see-though display but mirrors content from another device or offers a simplistic heads-up display.”

Headset market in flux

Meta again led the market in the first quarter in terms of share, while Apple’s recent entry into the market enabled it to capture the second position. ByteDance, Xreal, and HTC rounded out the top five, IDC said.

When online pre-sales of Apple’s Vision Pro AR/VR headsets began on Jan. 19 they sold out quickly, but as Computerworldnoted soon after, stable delivery dates could indicate limited demand for the $3,500 device.

Fast forward to April, and Apple said that it had cut Vision Pro production due to low demand, according to Ming-Chi Kuo, an Apple analyst at TF International Securities.

Jitesh Ubrani, research manager for worldwide mobile device trackers at IDC, said that with mixed reality on the rise, “expect strictly virtual reality headsets to fade in the coming years as brands and developers devise new hardware and experiences to help users eventually transition to augmented reality further down the line. Meanwhile, extended reality displays are set to garner consumer attention as they offer a big screen experience today while incorporating AI and heads-up displays in the near future.”

Meanwhile, Ramon T. Llamas, research director with IDC’s augmented and virtual reality team, said that although ASPs for the overall market crested above the $1,000 mark, this is not representative of all products.

“ASPs for augmented reality (AR) headsets have almost always been above this price point, but ASPs for VR, MR, and ER headsets have typically been lower,” he said. “Apple’s Vision Pro drove ASPs higher for MR headsets, but the addition of lower-cost devices from Meta and HTC have kept those ASPs from going much higher. Meanwhile, there were many devices for VR and ER priced below $500.”

Return to growth

Looking ahead, Llamas said that IDC is anticipating ASP erosion across all products: “Because the overall market is still in its early stages with more expensive first- and second-generation devices, prices will be high even as early adopters buy them. In order to reach scale in the mass market, vendors will need to reduce prices on later and upcoming devices.”

IDC is forecasting that “headset shipments will return to growth later this year with volume growing 7.5% over 2023. Newer headsets and lower price points will help with the turnaround expected later this year. Beyond that, headset shipment volume is expected to see a compound annual growth rate (CAGR) of 43.9% from 2024–2028.”

(Editor’s note: This column originally appeared on Computerworld Sweden on June 14, 2024.)

Just as everyone expected, and almost demanded, Apple finally started talking about artificial intelligence — in its own way, of course. The big keynote at WWDC on Monday might not have been the AI ​​event many had thought was coming. For example, the deal with Open AI, where Chat GPT will be used as an extension of Apple devices’ own AI capabilities, was negotiated in a matter of minutes.

Apple appears to be approaching AI with caution. Cautious, you might call it, but I actually think this strategy is the right one, and it aligns with what I called for earlier: AI that integrates seamlessly and easily into solutions we already know and use.

Apple Intelligence (of course Apple’s AI has been trademarked) is not a special app, or a special assistant or a “Copilot.” These are small, clever features, built on small, specialized models, sprinkled throughout the software. In Siri, in the photo app, as a writing aid, and so on, all in a seemingly non-intrusive way — an extra function, or help, that is there, if you want it.

The latter is important because it bothers me enormously is when AI is shoved down one’s throat. Just because an AI feature exists, maybe I don’t want to use it? No one but I knows what tasks I’m better at than AI, and it obviously varies from person to person.

For example, I am very good at writing and processing text. I definitely don’t want any AI getting in there (I even turn off the spell check in Word). On the other hand, sitting with transcriptions and translations is boring as hell, so I’m happy to take help there.

I’m a decent hobby photographer and don’t need an AI to make my photos “better” unsolicited. However, it can be fun or effective to take AI help to remove some ugly detail, play with the depth of field, or expose subjects.

I’m also a frequent user of chat, both privately and at work, but I think it feels a bit dirty to click on the suggested answers in Microsoft Teams chat (“Great”, “That sounds good.”) because it feels quite disrespectful to the person I’m communicating with.

BAbove all, I am seriously uninterested in Google’s new “AI Overviews,” which have now been rolled out, starting in the US. The AI ​​function in Google’s search engine takes the liberty of using AI to try to guess what you are looking for — and answer it.

I’m extremely good at Googling; it’s a skill I’ve developed over many years. And when I do research with the help of Google, it’s not one answer I’m looking for, but a balanced assessment that I make based on the information I google, thank you very much. Even if Google’s AI in the future gives “correct” answers instead of suggesting to glue the cheese on pizza, that’s just not what I want to use a search engine for.

So that’s why I think Apple is right here. It is through these kinds of simple, friendly and optional functions that do not require advanced “prompt engineering” that the masses will be introduced to and actually use AI tools. Because even though it might sound like it sometimes, most people don’t use Chat GPT at all.

Now Apple has the luxury, if you call it that, of not having to position itself as an “AI company” as a number of other tech giants want to do, although there has been pressure from investors to start delivering in this area. Apple sells mobile phones (and other hardware, but mainly phones). Therefore, it can be worthwhile to focus more on data protection and privacy, and on introducing features at a pace and in a way that makes mobile phone buyers see value in their presence.

Moreover, Apple isn’t charging extra for it, as most others do. Of course, Apple Intelligence is so far only available on the iPhone 15 Pro and Pro Max (and Mac computers with M-chip). And, presumably, that sprinkling of AI isn’t so sparkling yet as to warrant an immediate upgrade for most people.

But even if this particular iteration of Apple Intelligence will not become everyone’s everyday AI — anymore than the first iPhone became everyone’s smartphone — I believe, this is the way development will go. AI is fundamentally a commodity, a general-purpose technology.

It’s a feature, not a product.

This column is taken from CS Veckobrev, a personal newsletter with reading tips, link tips and analysis sent directly from Computerworld Sweden‘s editor-in-chief, Marcus Jerräng. Do you also want the newsletter on Fridays? Sign up for a free subscription here.