Generative AI capabilities — or at least the perceptions surrounding them — are rewriting almost all of IT. But with generative AI ROI proving elusive in IT circles, is AI truly going to impact near-term revenue? With mobile devices, analyst group IDC thinks it very well might.

“Premium markets are starting to embrace gen AI smartphones, generating excitement and renewed interest in the industry,” IDC said in a statement accompanying the latest edition of its Worldwide Quarterly Mobile Phone Tracker study.

This forecasts that Android phone sales will grow 7.1% in 2024, while sales of phones running iOS will stagnate, growing just 0.8%.

IDC forecast that generative AI-capable smartphones will capture 18% share of the total market by the end of 2024, with most flagship phones including some on-device generative AI features. However, the average selling price for smartphones with generative AI capabilities will be more than double the cost of those without, it said.

Adoption of AI-capable smartphones is of relevance for enterprises wishing either to deploy the technology internally, or push AI capable applications to customers.

“There is a potential upside to the iOS forecast with a lot depending on how well the demonstrated gen AI use cases play out in the upcoming iPhone 16 launch and how soon Apple can establish local AI partnerships in China,” IDC senior research director Nabila Popal said in the statement.

The use cases Popal is referring to involve Apple Intelligence, which is Apple’s attempt to deliver generative AI functionality on-device instead of via the cloud. Apple is discovering that such a delivery is complicated; the company’s September launch event for new mobile devices may focus more on what capabilities will materialize later rather than what can be delivered on launch day.  

If buyers are disappointed by the first generation’s weaker-than-hoped-for functionality, that might depress sales of second-generation products — even if the functionality with second-generation is far superior: The perception game can have a nasty boomerang effect.

Generative AI is not yet delivering growth for mobile but it is “creating excitement,” Popal told Computerworld.

She sees a key driver of generative AI in mobile being not the capabilities — at least not initially — as much as the perception of better data privacy.

That perception comes from the argument that much of the new data from device-resident AI will stay on the device and not move to the cloud.

“I don’t want all of my data to be on the cloud” given that the new gen AI functionality will be accessing so much personal information, she said. “It will be context aware, history aware, aware of my schedule” and aware of who the user is visiting and their planned visits.

The privacy reality, though, is that some — and potentially all — of that data may wind up in the cloud anyway given backup choices.

“Some data will still be stored on the cloud,” depending on user setting choices, Popal said. “That’s where the consumer consent comes in.”

The bulk of the potential excitement comes from the theoretical ability of on-device generative AI to break down the barriers between applications. Instead of the user looking up information from the larger number of installed apps, the device could answer the questions directly as it will interrogate the apps seamlessly in the background. 

The mobile model used to talk about “there’s an app for that,” Popal said, but as the number of apps soared, there was more of an effort burden placed on users. In theory, she argued, on-device AI could change that.

“Gen AI smartphones will completely change the way we interact with our phones,” she said. “Whether it’s processing a return, buying a ticket, booking a hotel or making sure it is the best deal, it will make human suggestions such as ‘If you change that date one more day, you’ll save $500.’” 

But — lest buyers get too excited already — such capabilities may take years to appear, as software vendors will need to adapt to new APIs as well as wait for their customers to buy AI-capable phones.

A new AI feature called ” Take notes for me ” will be soon be rolled out to Google Meet users. It’s a feature that creates automatic notes using Google’s AI ​​tool.

When the meeting is over, participants are given the opportunity to download a Google document with a transcrip of what was said. The function can also be used by meeting participants who entered the conversation late to find out what was said earlier.

Initially, “Take notes for me” will only work with conversations in English.

Meta has signed an agreement with Sage Geosystems to purchase up to 150 megawatts of geothermal energy to power its US data centers. The first phase of the project is expected to be operational in 2027, according to Reuters. And, according to Meta, it will significantly increase the use of geothermal energy in the United States.

Geothermal energy is renewable energy that uses the heat stored in the earth’s crust, which originates in the earth’s formation and decay of radioactive elements.

The new agreement is part of Meta’s ongoing efforts to meet the increasing demand for energy that has arisen with the company’s major investment in generative AI (genAI) technology, which is very energy-intensive. The financial terms of the agreement were not released.

Apple on Monday confirmed its next big public event will be Sept. 9, when it is expected to introduce the new iPhone 16 range, new Apple Watch models, and AirPods 4, with a focus on artificial intelligence (AI) hinted at by the event slogan, “It’s Glowtime.”

The event will be live-streamed as in past years across YouTube, Apple TV, Apple’s Developer app, and via the company website. The event begins at 10 a.m. PT/ 1 p.m. ET.

Why Glowtime?

When the company introduced Apple Intelligence at WWDC in June, it unveiled an all-new Siri behavior. While for now Siri shows itself to be working on a query by displaying a round pulsing shape on the screen, the new Siri simply glows around the edges of the device. Most Apple commentators believe this is the reference made in the invitation to the event.

That attention to AI is expected to be the big focus this year. Apple knows it has a chance to build the world’s biggest active AI ecosystem across all the devices that comprise its platforms. That’s a real opportunity for the company. But given that competitors are also racing in the same direction, Apple knows it must make a big impression with September news.

While not all of the new Apple Intelligence features are expected to be available when iOS 18 ships, we do anticipate more will arrive later in fall, when new Macs and iPads are also expected to arrive.

What to expect from iPhone 16 range?

As previously discussed, the iPhone 16 range will include the iPhone 16, iPhone 16 Plus, iPhone 16 Pro, and iPhone 16 Pro Max. (Recent speculation has claimed a much thinner high-end model, and an entry level iPhone SE upgrade will debut next year.) All the new devices — including the latter — are expected to be capable of running Apple Intelligence, which otherwise requires an iPhone 15 or later.

Big changes in all the iPhones might include a new camera layout, Action and Capture buttons, along with a 5x telephoto zoom on the Pro range. The Pro devices will also get bigger, 6.3- and 6.9-in. displays and improved cameras; there is also talk of the Pro range gaining solid-state buttons, which will make them even more resilient and waterproof.

Apple might also put more memory into the iPhones to provide better support for Apple Intelligence, enabling more complex generative AI (genAI) features to take place on the device. Other tasks will run on Apple’s own secured servers, though users can choose third-party services such as OpenAI’s ChatGPT directly from their iPhone. New colors and a higher 2TB capacity option are also expected.

What’s important is that both the outgoing iPhone 15 range and incoming iPhone 16 models will immediately be the world’s first mass-market AI smartphones, equipped with powerful AI features to help get more work done. Not only that, but up to 25% of the iPhones shipped across the next 12 months might be made in India.

What to expect from the Apple Watch?

The Apple Watch Ultra 3 may once again be the most lusted-after smartwatch. While the internal components will change and the device should become much faster with a new processor, externally it will seem similar, reports claim. It might be available in more colors.

The Apple Watch Series 10 will see more obvious improvements; slimmer, it will also host a bigger display. It is notable that this will be the 10th anniversary of the Apple Watch. Finally, the Apple Watch SE3 could be a lower-cost plastic model available in multiple colors and aimed at younger users, particularly as a way to keep in contact with kids in schools that ban use of smartphones. We could see new heath sensors introduced, but ongoing legal disputes might slow their introduction.

What about AirPods 4?

Though AirPods are not an enterprise product (yet), we know that wherever you have iPhones, you’ll often find AirPods. Now with USB-C charging, the new AirPods will become a range, with both lower-end and mid-range variants, leaving the AirPods Pro as the high-end choice. The AirPods 4 are expected to gain noise cancellation, which will be a great upgrade for many, particularly commuters, travellers, and workers seeking to focus at their desk.

What about iPads?

We’re not expecting iPad Pro news, though there is latespeculation Apple could introduce a new iPad mini and iPad model. These would be good upgrades, as they should host new chips, perhaps even the A17 processor used in the iPhone 15. Inclusion of new Apple processors is essential since the company will want to run Apple Intelligence on these devices.

While AI is the big marketing push, the inclusion of these new processors should make these entry-level iPads significantly more powerful than before. (The current 10^th generation iPad runs an A14 Bionic chip.)

There are implications for workplaces deploying iPads across their teams, as the upgraded devices will deliver much more bang per buck, (and computational cycles per watt).

What about Siri?

The Apple invitation’s reference to Siri hints that Apple will reveal even more AI features. That suggests the company hopes to make Siri a far more viable AI assistant than it is at present. Expect more personalization, smarter and more contextual intelligence, and predictive help. 

What about Europe?

Apple’s recent moves to tweak its approach to Europe’s DMA show the company has been in consultation with regulators there. While we can’t expect miracles, perhaps the company will be better able to explain its Apple Intelligence rollout plans for the region during the announcement.

More from Jonny Evans

• Jamf teams with Okta for enterprise-class simplicity
• How Apple can fight the tyranny of ‘choice’
• Microsoft Office apps circumvent Mac security

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

While the term “Open Source AI” has been used extensively in recent years, exactly what it means has never been clear — until now.

The Open Source Initiative (OSI) has released a proposed definition it hopes the tech world will accept. According to the new definition, it involves AI systems that consist of components that can be examined and studied. It must also be possible to freely modify the systems for any purpose and share them with other users, according to MIT Technology Review.

By that definition, the AI ​​models from Open AI, Anthropic, Google and Meta cannot be classified as “Open Source AI” because users are not allowed to do what they want with them.

In a stark warning to policymakers, Meta CEO Mark Zuckerberg and Spotify CEO Daniel Ek have emphasized that Europe could fall behind in the global race for AI leadership due to its fragmented and inconsistent regulatory landscape.

“With more open-source developers than America has, Europe is particularly well placed to make the most of this open-source AI wave,” Zuckerberg and Ek said in a joint statement.  “Yet its fragmented regulatory structure, riddled with inconsistent implementation, is hampering innovation and holding back developers.”

According to the two tech leaders, while AI holds the potential to transform industries and drive economic growth worldwide, Europe’s current regulatory environment could impede its ability to innovate and compete on the global stage.

“Instead of clear rules that inform and guide how companies do business across the continent, our industry faces overlapping regulations and inconsistent guidance on how to comply with them. Without urgent changes, European businesses, academics, and others risk missing out on the next wave of technology investment and economic growth opportunities,” the joint statement read

It can be recalled that in June, Meta had to postpone the launch of its Meta AI models in Europe as the Irish privacy regulator, Data Protection Commission (DPC), had asked it to delay harnessing Facebook and Instagram user’s data, to train its models.

“We’re disappointed by the request from the Irish Data Protection Commission (DPC), our lead regulator, on behalf of the European DPAs, to delay training our large language models (LLMs) using public content shared by adults on Facebook and Instagram — particularly since we incorporated regulatory feedback and the European DPAs have been informed since March,” Meta said in a statement in June. “This is a step backwards for European innovation, competition in AI development and further delays bringing the benefits of AI to people in Europe.”

The Irish DPC’s move came after NOYB, a Vienna-based digital rights advocacy group, complained to DPAs (data protection authorities) in 11 countries including Austria, Belgium, Spain, France, Germany, Greece, Italy, Netherlands, and Norway, to “immediately stop Meta’s abuse of personal data for AI.”

“Regulating against known harms is necessary, but pre-emptive regulation of theoretical harms for nascent technologies such as open-source AI will stifle innovation. Europe’s risk-averse, complex regulation could prevent it from capitalizing on the big bets that can translate into big rewards,” the joint statement by Meta and Spotify said referring to the Irish regulator’s decision.

Europe’s regulatory challenges and their global impact

Despite Europe’s strong tradition in open-source development, Zuckerberg and Ek argue that the region’s inconsistent regulatory framework is stifling innovation. They cite the uneven application of the EU’s General Data Protection Regulation (GDPR) as a prime example of regulatory uncertainty that is hindering progress.

“This landmark directive was meant to harmonize the use and flow of data, but instead EU privacy regulators are creating delays and uncertainty and are unable to agree among themselves on how the law should apply,” the two CEOs said in the statement.

Both the companies are of the opinion that these regulatory challenges are not just a European issue but have global implications. Restrictive policies in Europe could lead to a fragmented AI landscape, where innovations and technologies developed in other regions may not be easily integrated or adopted in Europe.

“Given the current regulatory uncertainty, Meta won’t be able to release upcoming models like Llama multimodal, which has the capability to understand images. That means European organizations won’t be able to get access to the latest open-source technology, and European citizens will be left with AI built for someone else,” Zuckerberg warned in the statement.

“The stark reality is that laws designed to increase European sovereignty and competitiveness are achieving the opposite,” the statement added.

Balancing innovation and regulation

To address these challenges, Zuckerberg and Ek called for a more streamlined and harmonized regulatory approach that balances the need for innovation with the necessity of protecting user rights and privacy. “With the right regulatory environment, combined with the right ambition and some of the world’s top AI talent, the EU would have a real chance of leading the next generation of tech innovation,” Zuckerberg stated.

“In short,” the joint statement pointed out, “Europe needs a new approach with clearer policies and more consistent enforcement. With the right regulatory environment, combined with the right ambition and some of the world’s top AI talent, the EU would have a real chance of leading the next generation of tech innovation.”

Google has reportedly put Character.ai founder Noam Shazeer in charge of its Gemini AI.

Shazeer worked in software engineering roles at Google for almost 18 years before leaving to found his own company. Now he’s returning to co-lead Google’s Gemini AI efforts alongside Jeff Dean and Orial Vinyals, The Information reported, quoting an internal memo.

The move follows the signing of a deal between Google and Character.ai earlier this month.

Google provided the startup with more funding in return for a non-exclusive license for its current LLM technology.

Shazeer and Daniel De Freitas — with whom he cofounded Character.ai in 2021 — were expected to join Google as part of the deal.

Other Character.ai employees were expected to stay with the company, though, expanding on the startup’s products, with the company’s general counsel, Dominic Perella, stepping in the role of interim CEO.

LexicaArt co-founder Sharif Shameem resurfaced an old anecdote about Shazeer’s hiring interview with Google during which he reportedly described a better way to correct spellings than the one Google was then using.

Shazeer is well respected in the field of AI, and has co-authored several research papers on the topic including important ones on LaMDA, PaLM, and ST-MoE.

Both LaMDA and PaLM have been heavily used by Google in its products and the company further expects to integrate Gemini into its products.

Google is not the only one courting expertise from startups. Rival cloud service providers such as Microsoft and AWS have also signed deals with smaller AI companies to either acquire the team behind their products, or their technology.

One of the biggest examples is Microsoft’s investment into OpenAI, following which Microsoft started integrating GPT-powered Copilots into all its products.

AWS and Oracle, too, have heavily invested into Anthropic and Cohere respectively in order to better their generative AI services on offer.