After weeks of speculation, ChatGPT creator OpenAI announced a new desktop version of ChatGPT and a user interface upgrade called GPT-4o that allows consumers to interact using text, voice, and visual prompts.

GPT-4o can recognize and respond to screenshots, photos, documents, or charts uploaded to it. The new GPT-4o model can also recognize facial expressions and information written by hand on paper. OpenAI said the improved model and accompanying chatbot can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, “which is similar to human response time in a conversation.”

The previous versions of GPT also had a conversational Voice Mode, but they had latencies of 2.8 seconds (in GPT-3.5) and 5.4 seconds (in GPT-4) on average.

GPT 4o now matches the performance of GPT-4 Turbo (released in November) on text in English and code, with significant improvement on text in non-English languages, while also being faster and 50% cheaper in the API version, according to OpenAI Chief Technology Officer Mira Murati.

“GPT-4o is especially better at vision and audio understanding compared to existing models,” OpenAI said in its announcement.

During an on-stage event, Murati said GPT-4o will also have new memory capabilities, giving it the ability to learn from previous conversations with users and add that to its answers.

Chirag Dekate, a Gartner vice president analyst, said that while he was impressed with OpenAI’s multimodal large language model (LLM), the company was clearly playing catch-up to competitors, in contrast to its earlier status as an industry leader in generative AI tech.

“You’re now starting to see GPT enter into the multimodal era,” Dekate said. “But they’re playing catch-up to where Google was three months ago when it announced Gemini 1.5, which is its native multimodal model with a one-million-token context window.”

Still, the capabilities demonstrated by GPT-4o and its accompanying ChatGPT chatbot are impressive for a natural language processing engine. It displayed a better conversational capability, where users can interrupt it and begin new or modified queries, and it is also versed in 50 languages. In one onstage live demonstration, the Voice Mode was able to translate back and forth between Murati speaking Italian and Barret Zoph, OpenAI’s head of post-training, speaking English.

During a live demonstration, Zoph also wrote out an algebraic equation on paper while ChatGPT watched through his phone’s camera lens. Zoph then asked the chatbot to talk him through the solution.

While the voice recognition and conversational interactions were extremely human-like, there were also noticeable glitches in the interactive bot where it cut out during conversations and picked things back up moments later.

The chatbot then was asked to tell a bedtime story. The presenters were able to interrupt the chatbot and have it add more emotion to its voice intonation and even change to a computer-like rendition of the story.

In another demo, Zoph brought up software code on his laptop screen and used ChatGPT 4o’s voice command app to have it evaluate the code, a weather charting app, and determine what it was. GPT-4o was then able to read the app’s chart and determine data points on it related to high and low temperatures.

Murati said GPT-4o text and images capabilities will be rolled out iteratively with extended “red team” access starting today.

Paying ChatGPT Plus users will have up to five times higher message limits. A new version of Voice Mode with GPT-4o will arrive in alpha in the coming weeks, Murati said.

Model developers can also now access GPT-4o in the API as a text and vision model. The new model is two times faster, half the price, and has five times higher rate limits compared to GPT-4 Turbo, Murati said.

“We plan to launch support for GPT-4o’s new audio and video capabilities to a small group of trusted partners in the API in the coming weeks,” she said.

OpenAI

What was not clear in OpenAI’s GPT-4o announcement, Dekate said, was the context size of the input window, which for GPT-4 is 128,000 tokens. “Context size helps define the accuracy of the model. The larger the context size, the more data you can input and the better outputs you get,” he said.

Google’s Gemini 1.5, for example, offers a one-million-token context window, making it the longest of any large-scale foundation model to date. Next in line is Anthropic’s Claude 2.1, which offers a context window with up to 200,000 tokens. Google’s larger context window translates into being able to fit an application’s entire code base for updates or upgrades by the genAI model; GPT-4 had the ability to accept only about 1,200 lines of code, Dekate said.

An OpenAI spokesperson said GPT-4o’s context window size remains at 128k.

Mistral also announced its LLaVA-NeXT multimodal model last week earlier this month. And Google is expected to make further Gemini 1.5 announcements at its Google I/O event tomorrow.

“I would argue in some sense that OpenAI is now playing catch-up to Meta, Google, and Mistral,” Dekate said.

Nathaniel Whittemore, CEO of AI training platform Superintelligent, called OpenAI’s announcement “the most divisive” he’d ever seen.

“Some feel like they’ve glimpsed the future; the vision from Her brought to real life. Others are left saying, ‘that’s it?’ he said in an email reply. “Part of this is about what this wasn’t: it wasn’t an announcement about GPT4.5 or GPT-5. There is so much attention on the state of the art horserace that for some, anything less than that was going to be a disappointment no matter what.”

Murati said OpenAI recognizes that GPT-4o will also present new opportunities for misuse of the real-time audio and visual recognition. She said the company will continue to work with various entities, including the government, the media, and the entertainment industry to try to address the security issues.

The previous version of ChatGPT (4.0) also had a Voice Mode that used three separate models: one model transcribes audio to text, another takes in text and outputs text, and a third model that converts that text back to audio. That model, Murati explained, can observe tone, multiple speakers, or background noises, but it can’t output laughter, singing, or express emotion. GPT-4o, however, uses a single end-to-end model across text, vision, and audio, meaning that all inputs and outputs are processed by the same neural network for more of a real-time experience.

“Because GPT-4o is our first model combining all of these modalities, we are still just scratching the surface of exploring what the model can do and its limitations,” Murati said. “Over next few weeks, we will continue iterative deployments to bring to you.”

Apple may have agreed to a deal to offer Chat GPT support within Siri on iPhones, according to the New York Times and Bloomberg. The deal might also eventually extend to other generative AI (GenAI) services, including Google Gemini, though no additional agreements are yet in place. This is just the latest in a rash of stories exploring the company’s AI plans.

It isn’t clear exactly how Apple will approach this. It’s possible the company will offer iPhone users the choice to use these services to replace its Siri voice assistant, and we’re likely to learn more about its implementation plans at WWDC in June. These will likely extend across its platforms.

How to secure Generation GenAI

Apple’s challenge will be to maintain the user privacy and security it prides itself in offering on its devices, while also supporting third-party chatbots. I do wonder if the company is investing in some form of OpenAI’s ChatGPT for Enterprise product, which provides additional privacy and security protection features.

The New York Times report confirms the company felt it missed the boat on genAI in the first instance, but also suggests the signal importance with which Apple regards this technology now. Among other things, it claims the company’s decision to close down its multi-billion dollar Apple Car project was made in order to transfer resources to the company’s in-house genAI development efforts. 

The report also describes some of the challenges secretive Apple has in attracting leading AI talent, noting that some new hires subsequently quit as they felt constrained by the need for secrecy. 

Setting up for a better Siri

For Siri, the main issue Apple execs hope to resolve through the deployment of ChatGPT is the capacity to deliver more accurate contextual understanding.

At present, Siri has no understanding of context, which means it just isn’t good at engaging in more complex tasks than stating the weather or turning music on. ChatGPT lets you define numerous factors to provide much more granular information. The company could also reach similar deals with other companies to offer their takes on genAI. (Apple is thought to have been in talks with both Baidu and Google in this regard.)

Bloomberg reports three primary strands to the company’s attempts:

• On-device genAI: Solutions that run on the device, no server required,
• Coud-powered AI: Cloud-based intelligence similar to that provided by ChatGPT.
• Chatbot services: A more intelligent, contextually-aware Siri. It’s possible Apple will send some or all Siri enquiries to a third party service, Bloomberg said.

What is at stake

The need to protect its iPhone kingdom is the primary motivation at Apple. The company apparently feels that the use of genAI, and intelligent agents powered by that technology, could eventually supplant both iOS and the App Store. “…it has the potential to become the primary operating system, displacing the iPhone’s iOS software,” wrote the New York Times, citing company insiders. 

Apple won’t be fully dependent on third-party AI, as it has its own solutions also in development. These include improving Siri’s ability to handle tasks it already manages, while extending its feature set with new tools the company can then introduce as being more private than other services. In part, this will be because some of the processing will take place, as we’ve anticipated, on the device itself. 

The report also confirms recent claims Apple is developing new silicon for use in data centers. The idea here will be to reduce server costs, improve energy efficiency and to bake privacy into the system in order to maintain user security.

Will WWDC be make or break for Cupertino?

It is important to note that the frequency of reports concerning Apple’s plans for AI is accelerating as we ramp toward the company’s big developer event, WWDC. This strongly suggests that Apple will introduce new components to its overall plan for AI at the show. 

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

Apple’s head of security engineering and architecture, Ivan Krstić, this week announced the publication of what should be essential reading for Apple admins and security pros — the newly updated Apple Platform Security guide. (Among other things, Krstić also leads Apple’s war against surveillance hackers.)

The latest update since 2022, the guide is currently being translated into local language versions, so it might not be available on your local Apple server. You can get it in American English directly from the US site, and you’ll know when you find it because the May 2024 publication date will be visible at the bottom of the front page. 

What is the Platform Security Guide?

“This documentation provides details about how security technology and features are implemented within Apple platforms. It also helps organizations combine Apple platform security technology and features with their own policies and procedures to meet their specific security needs,” Apple says in the introduction to the 210-page document. (It’s interesting to note that in 2019 the guide extended to 157-pages.)

Open it up and you’ll find updated information, along with the addition of new sections addressing several topics, including App Store, WidgetKit, and Lockdown Mode security. The latter doesn’t explain much we didn’t know already, but puts the protection into context and links to the most recently updated information concerning that mode. The document has also been brought up to speed with additional information concerning start-up security on the latest Apple Silicon devices and harmonizes links to the company’s security reporting pages.

I expect in the future it might further extend to sharing information pertaining to server chips from the company, if that plan turns out to be true.

What’s new in the Platform Security guide?

Some particular highlights include a better explanation of the company’s built-in malware protection system, XProtect, and a little added insight into how App Store security works. 

How XProtect works is to some extent a bit of a black box, but the latest iteration of the report does shed a little light on what’s happening:

“Should malware make its way onto a Mac, XProtect also includes technology to remediate infections. For example, it includes an engine that remediates infections based on updates automatically delivered from Apple (as part of automatic updates of system data files and security updates). This system removes malware upon receiving updated information, and it continues to periodically check for infections; however, XProtect doesn’t automatically restart the Mac. In addition, XProtect contains an advanced engine to detect unknown malware based on behavioral analysis. Information about malware detected by this engine, including what software was ultimately responsible for downloading it, is used to improve XProtect signatures and macOS security.”

As for App Store security, EU readers will note that this section hasn’t yet been updated to include what security Apple provides around purchases made from third party stores. That’s likely to make interesting reading once it does appear. But the document does explain the five different security processes that govern apps sold through the company’s own App Store. These include automated malware scans, human review, manual checks, user reviews, and processes for correction and removal of bad/scam apps.

Under the EU sideloading scheme, Apple will only be able to ensure malware scans and respond to user feedback; third-party app providers will deliver (and presumably in some cases, fail to deliver) the other security processes.

Who is the guide for?

This really is essential reading for anyone who wants to better understand Apple security. That means Apple admins as well as developers, security researchers, customers — anyone who really wants to get to grips with the information it offers.

Those already familiar with the document shouldn’t expect much; while there are some new sections (and dozens of sections have been updated), many of those changes are relatively small. (Some of the information about recently introduced security tools for Messages may be of interest, however.)

Given the scale and complexity of the Apple platform ecosystem, it seems likely some small tidbits of new information will be found. 

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

Nearly a year after rolling out a generative artificial intelligence (genAI) tool to help it process thousands of claims documents, global insurance claims management firm Sedgwick says the technology’s accuracy is nearly perfect.

Sedgwick, which operates in 80 countries, receives about 1.7 million pages of digital claims-related documents a day. The documents then go through an arduous vetting process by examiners who must decide whether they’re valid and how they should be handled.

In April 2023, Sedgwick unveiled a genAI tool called Sidekick to help with document summarization, data classification, and claims analysis. Sidekick uses OpenAI’s GPT-4, giving the company “an unlimited number of large language models to be created for varying purposes.”

In December, Computerworld spoke with the company’s global chief digital officer, Leah Cooper, about the challenges and purposes of the genAI rollout. At that time, Sedgwick’s Sidekick genAI technology had combed through 14,000 documents and was “shockingly good” at accurately spitting out summaries. Five months later, Cooper said more than 50,000 documents have been processed by Sidekick, and those documents have been evaluated by more than 1,000 examiners who reported a 98%-plus accuracy rate in the document summarizations.

Computerworld revisited the topic with Cooper to ask her what she and the company have learned about genAI and its capabilities for reducing workloads and increasing document processing efficiencies.

Tell me about the kinds of documents you had Sidekick evaluate. Are they all medical-related insurance documents or do they run the gamut? And how long are they? “Medical documents in the workers compensation space were the core focus for our initial pilot, but we have since expanded to other types of documents and photos for validation. The average length of the documents in the testing phase was six to seven pages, but some were much longer, ranging from 25 to 30 pages.”

Tell me a little about Sidekick, how you developed it using OpenAI’s GPT-4, and how it connects to your document management system? “We developed Sidekick so that we can leverage the best of what genAI has to offer, giving our claims professionals an advantage in their daily work.

“If we can drive efficiencies by taking the busywork out of claims administration, and allow our people to focus on taking care of our customers, we can transform not only the process but the experience of having a claim. Our first initiative was to summarize documents that are received in support of a claim. Those were the basics: deploying ChatGPT into the Sedgwick environment so that our data stayed secure, and then evaluating a first use case to see if genAI could be successfully implemented.  

“We are thrilled to say that we did that successfully, incorporating over 50,000 documents during our pilot phase of Sidekick. We’ve just wrapped up our second phase, where we integrated Sidekick technology into our proprietary claims admin systems. This was a huge part of the productivity driver in our business.

“Now, how do we make this tool more relevant to our business? How do we drive productivity, decrease resolution time, and shift from a tactical application to a strategic and conceptual one? This is where we are uniquely positioned: it’s not one tool, it’s several capabilities pulled together to create a scalable, rapidly deployable platform in a unique way. If we combine generative AI with 50 years of understanding and refining how the claims model works and a best-in-class data science program, Sedgwick will pivot into a business model that transforms the claims industry.

Last time we spoke, you told me about 500 employees were using Sidekick. Did that number remain the same and how did they use genAI in their work? “Now we have over 1,100 employees who have used this tool. Examiners are using this technology to summarize claims documents and expedite the entire process.”

What kind of feedback did employees using Sidekick give you – positive and negative (if both)? “Employees are actually asking us to make this product widely available more quickly. People who have used this solution are telling colleagues about the accuracy of summarizations and time saved on claims, fostering a culture of excitement around a new tool which hasn’t existed before.”

How did you determine the 98%+ accuracy rate? “In Sidekick’s pilot phase, we constantly asked for feedback from employees who were evaluating the output results of Sidekick. Examiners would be prompted to say whether the document was successfully summarized or whether something was missing.

“One key to defining a strong AI program is to set the expectation of outputs so users can understand what they are judging as part of this new process. By identifying what examiners are looking for and defining our output results, we were able to set a standard for what is deemed successful and what is not. 

“It’s incredibly important to obtain real feedback from the users who are ‘boots on the ground.’ Individuals who would normally create these summarizations manually were the ones who graded the AI. In the first few months, we did a lot of tweaking based on feedback and it took multiple iterations of prompt engineering to simulate what goes through an examiner’s mind.

“Once we nailed it down and were satisfied with initial testing scores, we rolled it out more widely to 1,100 employees, who ultimately scored Sidekick with the 98%+ accuracy rating. Business partner involvement in determining success is crucial to adoption of the technology. If the people that support these claims are not behind it, companies will not realize a successful engagement with technology.”

Can you explain the time savings and potential productivity increases genAI created? “This technology has created and will continue to learn efficiency gains throughout our organization. We’ve been trying to find a way to automate tasks associated with claims administration that are not complex for business operations, they’re just necessary steps in a process. From my perspective, we want to focus on how we can recognize areas that need lesser attention (e.g., the simpler claims) and allow our examiners to really focus on the other types of claims that could benefit from faster and more collaborative engagement with our customers. If we can take busywork out of claims that need minimal investigation and instead direct adjusters to a claim that needs more attention, clinical resources and time with customers, then we have changed the model for care throughout a difficult time in somebody’s life.”

You receive about 1.7 million claims-related documents every day, so the 50,000 documents handled by genAI is a relatively small percentage. What kind of methodology did you use to ensure these documents were an accurate representative of the whole? “During this rollout, we didn’t select only documents that had certain attributes. We wanted the big picture. In the end, we worked with clients who were anxious to collaborate with us on new technology-forward programs to analyze claims. We worked with clients directly to receive approval to use this new technology during our claims process and ran every document through the generative AI solution once they were on board. This gave us a solid exposure to every type of claim and document, ensuring that genAI was thoroughly vetted for every scenario.”

What are some things you’ve learned from the project? “When we first integrated genAI into our operations, we had to learn quick about how and when to best apply it. GenAI technology is rapidly evolving, it seems like on a daily basis. While this is transformative, it also means that we have a perpetual learning curve and challenge to understand the best application of genAI. It gives us an opportunity to work in the most agile environment imaginable: this is a very exciting, and overwhelming, time for leadership. 

“A great deal of articles and media around this topic have compared genAI to the introduction of the internet, and while they’re a bit different, there are some major similarities. Sedgwick and companies around the world are investing in these tech tools, but the companies that will see the most success are the ones that think critically about how to best leverage this technology. By identifying how genAI best fits into operational models — which can be challenging as much of the tech is developed in a vacuum — it is vital to identify the best use-cases for each tool. As these challenges and opportunities continue, we’re excited to see what new opportunities arise as innovation continues.”

Did you encounter anything unexpected throughout this trial of Sidekick? “We encountered a number of positive takeaways from this trial of Sidekick. Namely, the reliability of the genAI products currently available were much higher than we expected.

“Initial accuracy rates were staggeringly high, and they have proven to be incredibly useful for our claims management teams. In the past, tech tools ‘out of the box’ have not been this effective. That initial success highlights the rapid innovation, which will continue, of artificial intelligence right now and the current and potential use-cases for it across business models and operations.”

What’s the next step in your genAI journey? Will you be instituting a larger rollout, or are you considering other areas of the business for its use and, if so, what are they? “The next steps … involve a focus on transforming workflows through the combination of new tech tools, along with data science, decision engines, and dynamic API outreaches. This combination of tools into a new platform will enable low-touch automation on simple claims like never before. Our operational model and understanding of the industry climate has already set the stage for our ‘must haves.’ We understand that better than anybody out there. 

“However, our latest genAI release lets us recognize, ‘OK, we just learned this from new data or supporting documents. What does that mean to the claim?’ That’s where data science steps in: through our years of best-in-class operations and collaboration with clients, we have a data set that lets us know what happens next. Taking the information that we learned from generative AI, and combining that then with the analytical AI of predictive modeling, we can drive the advancement of a claim and provide prescriptive recommendations to our claims examiners. 

“And that’s an important point here. The goal is never to replace the critical thinking and judgment calls that our people do so well. It’s to inform them with relevant, rapid data so that they can make those decisions even better. To put it simply, we will be able to say, ‘This claim has taught us this, so expect that.’

“We can address the next step in claim lifecycle, and we can think further out to find an optimal path for resolution. Claim resolution time will recede, early adopters will reap benefits of technology enablement more rapidly, and the experience of those who we serve will improve. But this triangle that brings together genAI, data science, and 50 years of knowledge has uniquely positioned us to make that claims model more intelligent and more situation specific than ever before.”

Do you have any tips or recommendations for others considering rolling out genAI? “The AI sector has been funded and developed beyond the knowledge and understanding of how to use these tools, and companies are struggling to find a way to integrate this technology with legacy processes. It’s important to focus on meaningful ways to transform the way companies do business by adding in resources that help shape judgment calls without removing a human from the claims experience. 

“I would say that facilitating a strong partnership between tech and operations is key to figuring out this genAI journey. You have to approach the work as if no processes are sacred, and companies and employees can’t be afraid to leverage AI to find new innovations and efficiencies.

“And for a last tip, I like to call this ‘digital triage,’ in that there cannot be the assumption that there is a blanket application out there that will be useful across an entire business model. Take our work, for example, with complex claims, it is necessary for a human to be there to partner with the person submitting their claim, and by leveraging a tech solution such as genAI, the groundwork can be laid for humans to focus on the most important aspects of the process.”

Mandatory return-to-office policies appear to be pushing workers at major tech companies away from their employers, with measurable effects on how willing people are to stay at companies that require in-person attendance.

A study released this week by researchers at the University of Michigan and University of Chicago found that three large US tech companies — Microsoft, Apple and SpaceX — saw substantially increased attrition, particularly of their more senior personnel, when they implemented strict return-to-work policies in the wake of the COVID-19 pandemic.

While many efforts to study remote work and its effects on the economy have been based on survey data, the authors of this study used publicly available resume data, rather than self-reported preferences, to track the actual effects of back-to-work policies

The three companies were chosen, according to the researchers, because they were among the first to implement return-to-office mandates as the pandemic eased in 2022, and because of their critical importance to the technology sector.

“We estimate nearly identical effects for all three companies despite their markedly different corporate culture and product gamut, suggesting the effects are driven by common underlying dynamics,” the report said.

One of the most striking findings, according to David Van Dijcke, one of the study’s authors and a Ph.D candidate in economics at the University of Michigan, was that workers in more senior positions were more likely than their juniors to leave a job rather than go back to the office.

“You might expect something different, where younger or more junior employees have matured in or started their first jobs in a remote environment,” he said. “But we didn’t find anything like that.”

The study also found a tight correlation between the rigor of a specific return-to-office policy and its effects on workers, Van Dijcke said. Apple’s one day per week policy produced the smallest changes to its workforce, causing about a 4% decrease in senior employees as a share of the overall pool. SpaceX’s full-time in-house requirement, by contrast, led to a larger than 15% decrease.

Van Dijcke, who is also employed at the risk analytics division of Ipsos Public Affairs, said that there were several possible reasons more senior tech workers might leave. Another study, he noted, tracked two offices of the same company, located mere blocks apart, and found that employees there got different benefits from working remotely and working in-office.

“They found that senior software engineers were more productive when they were not close to their co-workers, which I guess makes sense, right?” he said. “They’re skilled at what they do and benefit from fewer distractions. Whereas if you’re more junior you might benefit from distractions that give you valuable feedback.”Recent surveys have tracked closely with Van Dijcke and his colleagues’ findings, including data published today by Gartner finding that one in three executives would quit over a return-to-office mandate, compared to less than 20% of non-executive workers.

Nevertheless, Gartner said that in-office requirements are getting more strict across the technology sector. Dell, for instance, has begun to issue employees color-coded “grades” based on their attendance.

For years now, Apple has been as much a marketing company as a technology company. But after Apple CEO Tim Cook introduced the iPad Pro ad this week, you have to ask whether Apple has lost its marketing mojo. 

The ad features — if that’s the right word — musical instruments, artist tools, toys, and games being crushed by a huge hydraulic press and turned into a new iPad. All this happens to the musical background of Sonny and Cher’s 1971 hit, All I Ever Need Is You. 

Ah, no, Apple, we need far more than just you.

I mean, seriously.  How was this ad ever made? Who greenlighted it? And how in the world did it ever make it to the public? Did no one even look at it?

It’s not just me. Many other people hate — really hate — the ad, based on the groundswell of criticism that erupted online this week. 

At a time when artificial intelligence (AI) fans are pushing out artists, writers, and musicians, it’s offensive for Apple to imply we can do without all those old analog creative types. (I’m reminded of how James McNerney, former Boeing CEO, would call his company’s senior engineers  “phenomenally talented a**holes.” If Apple ends up shoving out its creative people the way Boeing did its engineers, future Apple products might be as “good” as the Boeing 737 Max 9, and the 787 Dreamliner.)

No one will die from a bad Apple, but it’s a red flag whenever a company mistreats its top people. It’s a slap in the face to all of Apple’s in-house creatives, as well as the company’s many loyal creative users. 

The timing could not have been worse. People, especially creative pros, are understandably nervous about their work being stolen and losing their jobs to soulless AI algorithms and apps. This ad is that very fear crystallized into 60 seconds of video. 

This is not the kind of “Think Different,” people want to see from Apple. Indeed, this is the opposite of what people want. As Michael J. Miraflor, chief branding officer for venture capital company Hannah Grey VC, tweeted, “I don’t think I’ve ever seen a single commercial offend and turn off a core customer base as much as this iPad spot.“

James Cook, marketing director at venture capitalist firm Molten Ventures, summed it well on Xitter: 

Apple’s new “Crush” ad (let’s call it “2024”) is a visual & metaphorical bookend to the 1984 ad.

1984: Monochrome, conformist, industrial world exploded by colourful, vibrant human.

2024: Colourful, vibrant humanity is crushed by monochrome, conformist industrial press.

Exactly so. 

Now, Cult of Mac members might argue we just don’t get it. That just tells me that Apple’s marketing policies, started by Steve Jobs with his personal reality distortion field, have been remarkably persistent. But that was then; this is now.

Seeing objects like pianos and guitars that many love being smashed is a slap in the face and it’s painful to watch. This is how Apple can lose its marketing mojo.

And if you don’t think Apple could lose its popularity due to an ad, think again. Ask Coca-Cola how  its 1985 New Coke campaign worked out. Hint: Many of you have never had a New Coke in your life. More recently, Peleton saw its stock price drop by 15% after its ad campaign showing a woman being thankful to her husband for getting an the exercise bike so she could become prettier. Yeah, that went over well.

Remarkably, the same concept was done much better by eBay, of all companies, in 2015, eBay produced an animated GIF of the evolution of the desk where desktop items from photos to calculators to calendars transform into icons on an ever-changing computer screen. 

Unlike the Apple ad, this was fun, and it made the same point.

Even more annoying is that Apple could have used the same footage and music to make a brilliant ad. That’s what Reza Sixo Safai did on Xitter. Safai simply reversed the video, so from an iPad Pro, the crushed wreckage is restored to all its wonderful, delightful glory. 

The moral of this story is that no company can ever be so successful that it can’t blow it. If Apple is as smart as its fans think it is, it will dump this ad as soon as possible and try to find another way to get people excited about buying iPads. 

This was not the way. 

Generative AI (genAI) tools are becoming more common in the workplace, but business leaders are concerned that their organizations lack a strategy to deploy the technology across their workforce.

That’s according to a Microsoft survey of 31,000 employees in 31 countries, published in the company’s annual Work Trend Index report. 

The survey indicates strong demand from employee for access to genAI tools. Three quarters of respondents use the tools in their jobs, the report claims, double the usage of just six months ago. Employees say AI saves time, enables them to focus on more important tasks, allows for more creativity, and lets them enjoy work more. And more than three-quarters (78%) of office workers use their own AI tools —  a phenomenon known as bring your own AI (BYOAI).

Business leaders also see potential in the technology, with 79% of leaders surveyed believing AI use is needed for their organization to remain competitive. 

Microsoft itself has claimed several large-scale deployments of its own Copilot genAI assistant: Amgen, BP, and Koch Industries are among the enterprises that have purchased over 10,000 “seats” of Microsoft 365 Copilot, CEO Satya Nadella said during the company’s recently quarterly financial earnings call. 

But not all large businesses are keen to dive in quickly, however. The survey found that 60% of leaders believe their organization’s leadership lacks the vision to roll out AI across their workforce. 

“While leaders agree using AI is a business imperative, and many say they won’t even hire someone without AI skills, they also believe that their companies lack a vision and plan to implement AI broadly; they’re stuck in AI inertia,” Colette Stallbaumer, general manager of Copilot and Cofounder of Work Lab at Microsoft, said in a pre-recorded briefing.

“We’ve come to the hard part of any tech disruption, moving from experimentation to business transformation,” Stallbaumer said.

While there’s clear interest in AI’s potential, many businesses are proceeding with caution with major deployments, say analysts.

“Most organizations are interested in testing and deployment, but they are unsure where and how to get the most return,” said Carolina Milanesi, president and principal analyst at Creative Strategies.  

Security is among the biggest concerns, said Milanesi, “and until that is figured out, it is easier for organizations to shut access down.”  

As companies start to deploy AI, IT teams face significant demands, said Josh Bersin, founder and CEO of The Josh Bersin Company. Deploying genAI tools puts the onus on IT staff to ensure data quality and security standards are in place, as well as “getting up to speed on the European AI Act, implementing governance, and helping to standardize on vendors and tools, if possible,” he said. 

With all this groundwork required, it’s likely to take a year or more for businesses to develop a comprehensive strategy around genAI, said Bersin.

Does GenAI generate business value?

Another sticking point is determining value and ensuring a return on investment when investing in AI.

AI takes many forms, but genAI is the focus of most newer AI initiatives within organizations, according to a recent Gartner survey. The most common way employees interact with the technology is when it is embedded into existing productivity and line-of-business apps (34% of respondents), such as Microsoft’s 365 Copilot, Adobe Firefly, and many others. 

Other ways to interact with genAI, include prompt engineering (25%), training bespoke genAI models (21%), or using standalone generative AI tools, such as OpenAI’s ChatGPT or Google’s Gemini (19%).

These genAI features don’t come cheap. In most cases, digital work app vendors charge an additional fee for generative AI features within their paid products. This can be as much as an extra $30 per user each month for Microsoft and Google’s business-focused AI assistants, while those with a more limited focus can charge less, such as Slack AI, which costs a not-insignificant additional $10 per user each month. 

Alongside the challenges in measuring the impact of genAI, spending significant sums in training up employees can also be seen as a risk. 

Demonstrating the value of AI projects is cited as the biggest obstacle to AI adoption, according to the Gartner survey. “As organizations scale AI, they need to consider the total cost of ownership of their projects, as well as the wide spectrum of benefits beyond productivity improvement,” Leinar Ramos, senior director analyst at Gartner, said in a statement.

Microsoft’s survey paints a similar picture. The majority of leaders (59%) are unsure of their organization’s ability to quantify any productivity gains from employee use of AI.

“Cost is really where organizations are getting stuck,” Milanesi said, with companies unsure of what returns they can expect when deploying generative AI. 

Bersin said many organizations have seen productivity improvements in early trials of genAI tools, but a shift to broader, company-wide deployments requires greater consideration of the value it can deliver. “When it comes to massive purchases across the enterprise, I am sure there will be lots of discussions about ROI, because these tools are expensive,” he said.

In its report, Microsoft cited a six-month, randomized control trial of 60 Copilot customers involving 3,000 workers. Results included an 11% reduction in emails read and 4% less time interacting with them, as well as a 10% increase in the number of documents edited in Word, Excel, and PowerPoint. The impact on the number of meetings was less clear; some companies saw an increase, others, a drop. 

There’s a tendency to focus on time saved when assessing genAI tools, said Milanesi, but that might not be the best approach. The real value is in improved quality of work and increased worker satisfaction, she said.This drives “better engagement at work and, in turn, better work,” she said.

Where a tool boosts employee engagement, cost becomes less of a consideration. “Think about the cost of a worker quitting, or someone staying in the job and not being engaged,” she said.

It might be that some AI tools are better suited to certain job roles than others. “The question for any leader is to identify what level of AI is right for the talent. Like for PCs, not everybody needs the top of the line,” said Milanesi. 

Apple’s Let Loose event saw the company introduce an iPad Pro equipped with a next-generation M4 chip, an iPad Air running an M2, a new Apple Pencil Pro, and powerful updates to the company’s pro apps on iPads.

There’s a lot to like about all these announcements. The iPad Pros are thin — at 5.1mm, marginally thicker than the USB-C slot they use. They are also highly performant, with better cameras and improvements across the board. The M4 chip shows Apple pushing far ahead of the entire tablet (and PC) industry, as do the advanced Ultra Retina XDR display and LiDAR capable pro camera. (Apple CEO Tim Cook calls the display, “the most advanced display we’ve ever produced.”)

But it’s not just pulling ahead in the tablet business. 

Setting the stage for AI with better chips

Apple made no secret of the built-in AI capacity ready to be pulled out of these things. It doesn’t just want video editors and musicians getting creative with iPads — it also wants data scientists and machine-learning developers using these oh-so-portable products. The fact that the M4 chip in the Pro is capable of 38 trillion operations a second is just icing on the cake.

These tablets are built to eat up all of the on-device computing you can throw at them. That’s great for the iPad and spells an even brighter future for the Mac. How else can you see Apple’s boast that the M4 chip means the high-end iPads are now already “more powerful than any AI PC today.” In comparison, the company claimed the chip delivers 60 times the performance of the A11 chip of seven years ago.

“The new iPad Pro with M4 is a great example of how building best-in-class custom silicon enables breakthrough products,” said Johny Srouji, Apple’s senior vice president of hardware technologies.

Apple has also introduced machine-learning accelerators to improve features you use a lot, including FaceID, QuickType suggestions, photo categorization and more. I’ve a feeling that power is going to mean so much more later this year when the company introduces new AI features in its operating systems. Like so many previous Apple releases, what these things can do is only going to show itself over time.

The M4 chip: Performance excellence

The new processor hosts the fastest Neural Engine Apple has made yet. It also pushes an even wider lead between Apple’s silicon and promised competing chips — even the best of which continue to struggle to match Apple’s still-new M2 processor. What matters is that the M2 is only just two years old, meaning Apple now has a road map for processor innovation that should give anyone turning to its platforms confidence in their longevity.

It means if you move to Apple, you’ll get cutting-edge performance that will last years, and you won’t need to migrate anywhere because the manufacturer goes bust or disappears. It also means Apple remains at least two generations of nearest mobile chip rival, Qualcomm.

iPad Air purchasers get that M2 chip, which was used in previous iPad Pro models. That processor is a 5nm chip.

Some stats:

• The M4 is 1.5 times as fast as that two-year-old M2 processor. 
• The M4 also provides up to 10 CPU cores — four performance and six efficiency cores — and it supports hardware accelerated AV1, H.264, HEVC, and ProRes. 
• The GPU also gains 10 cores, hardware-accelerated ray tracing and mesh shading.
• The new on-chip display engine supports both OLED and ProMotion with 10Hz-120Hz refresh rates.
• The M4 delivers the same performance as the M2, but demands half as much power.

The first Geekbench tests claim 9,234 points on the Core ML Neural Engine Inference test for the 10-core version of the chip. More robust test results will no doubt appear once the device ships May 15.

Oh, and Apple strongly hinted that M4 Macs will arrive later this year.

About the new iPads

It is interesting to explore the different iPad Pro models, as you’ll find huge differences between them. For example, the 256GB/512GB models carry 9-core CPUs, while the 1TB/2TB storage models have 10-core processors. Memory on the larger capacity models is double that of the lower end.

You can argue that iPads have now become sufficiently mature that Apple can make very different configurations, in a similar way to the different Mac SKU’s. That also means if you are planning to invest in these tablets for you or your teams, do check which models are used when considering reviews of any apps you happen to need.

You also need to know that all Apple’s cellular iPads are eSIM only in all markets. And Apple no longer provides a charger in the box for iPads sold in the European Union or UK.

Apple

iPad Pro

No doubt the iPad Pro is the star of Apple’s tablet dance. That new M4 chip is built on a second-generation 3nm process tech, which means it delivers desktop performance with mobile battery life. It’s a good sign of what’s to come from Apple in the next 12 months: supercomputer performance distributed across a wide range of platforms. This will be the platform of choice for personal genAI — at least, I think that’s what Apple hopes.

The chip isn’t the only thing that’s small. The device is thinner than an iPod nano (remember them? I keep one on my desk). How thin?  5.3mm on the 11-in. model and 5.1mm on the 13-in. models. You also get less weight and a smaller battery which still delivers up to 10 hours of active life, thanks to the increased power efficiency of the processor.

The new display

The iPad Air display remains the same as the previous model. The iPad Pro, however, uses what Apple is calling a Tandem OLED display. It consists of two panels stacked one above the other to deliver better color accuracy, brightness, and longevity. Anecdotal reports are extremely enthusiastic about the screen quality, the best in any tablet today. Data-wise, expect 1,000 nits of peak full-screen brightness, and 1,600 nits of peak HDR.

Apple had been expected to introduce the screen, but what is surprising many is the cost. Display Supply Chain Consultants (DSCC) founder and CEO Ross Young had said he thought the move to this kind of display combined with the new M4 chip would mean a much higher product price — he had estimated prices would be 50% higher. (You can spend more if you want; for $100 Apple will apply a nano-texture on the display surface to combat glare.)

Additional details include:

• 12MP rear and front cameras, with the front camera at center of screen edge. Apple has silently removed the 12MP ultra-wide camera it fielded in these iPads before.
• Prices start at $999 for the 11-in. and $1,299 for the 13-in., available for pre-order today and hitting stores May 15.
• Two colors: silver or black.
• Support for the new Apple Pencil Pro.
• Dimensions: 11-in. — 249.7mm-x-177.5mm-x-5.3mm; and for the 13-in. — 281.6mm-x-215.5mm-x-5.1mm. 
• Weight: 11-in: 444g or 446g with cellular. (roughly 0.98 pounds). 13-in: 578g or 582g with cellular. (roughly 1.28 pounds).

More specifications are  available on Apple’s website.+

About the iPad Air

Apple confirmed speculation announcing both 13- and 11-inch models of the iPad Air, now with M2 processors and the same LCD processor as the previous model. The introduction of M2 chips to the iPad Air is a big step up for that tablet, but it’s clearer now than ever where the investment is taking place – iPad Pros now weigh even less than their lower priced “Air” sibling (579g versus 617g). The 6.1mm Air is also thicker than the Pro.

Additional details include:

• 12MP rear and front cameras, with front camera at center of screen edge.
• The iPad Air is available in four colors starting at $599 for the 11-inch model and $799 for the new 13-inch configuration.
• Brightness ranges between 500-600nits.
• Supports new Apple Pencil Pro.
• 13-inch: Dimensions: 280.6mm-x-214.9mm-x-6.1mm.
• 13-inch: Weight: 617g or 618g with cellular. (c.1.36 pounds).
• More specifications are available on Apple’s website.

Buried in the tidal wave of information, Apple also lowered the price of its entry level iPad to $349, $100 less than the original price. I’m not certain, but I’d like to know if this is the only thing in the world that’s gone down in price, given everything else has seen costs increase. The ninth-generation iPad previously sold at $349 has been removed from sale, taking with it the last Apple tablet sold with either a headphone jack or TouchID button. 

The Apple Pencil goes Pro

The Apple Pencil Pro ($129) deserves its own mention. This is effectively a computer in its own right, with squeeze controls, haptic feedback, and a range of gyroscopically handled tools that make the device both precise and intuitive to use. It also includes Find My, so you won’t lose it, and charges while magnetically connected to the tablet.

The Magic Keyboard, too, has seen a really significant update. Not only does it deliver an aluminum palm rest, but also a larger trackpad that provides haptic feedback and even hosts a row of function buttons. Available in silver and space black, the addition of the keyboard means using the iPad will feel far more like the experience you expect using a MacBook. At (from) $299, they aren’t cheap, but with the slim iPad they promise a mobile computing experience second to few.

Finally, some environmental achievements. Apple claims the iPad Pro comprises over 20% recycled or renewable content, including 100% recycled aluminium and 99% recycled rare earth elements. All the tin, gold, and copper used on printed circuit boards are also recycled. You can read the product environmental report here.

Who are these iPads for?

Apple continues to lean into creative markets with the iPad range. To support this, it introduced powerful updates to both its Final Cut Pro and Logic Pro lines, introducing new tools and features and integrating even better support for multiple devices. You can now use one or more iPhones and iPods to capture multicam footage, for example. Apple also announced new AI features, enabling artists to get more done faster and better.

All the same, I think Apple is missing something here. We know that iPads are seeing increasing use across the enterprise. Retail employees, frontline customer services, aircraft cabin crew (and pilots), field service engineers, and huge numbers of C-suite executives already use iPads at work; I think Apple should work a little harder to celebrate that.

It does have some information about this; a search of its enterprise website yields case studies explaining use of iPads at United Airlines  and across the Tokyo Metro, for example. However, I think that by focusing so much on creative uses, Apple is missing the chance during these launches to celebrate some of the less colorful but still incredibly valuable ways in which its powerful mobile devices are used. These iPads aren’t just for creatives and consumers, they are suitable for a growing array of uses across business and commerce, too.

What analysts said

Carolina Milanesi at Creative Strategies said: “It seems that Apple is positioning the silicon as the real way they are competing in the AI world. It’s really about enabling developers more so than it is about Apple bringing to life a new Siri or services that are AI-driven. This could all change at WWDC, but we always need to go back and remember the business model. While for Microsoft or Google everything goes back to the cloud, for Apple, everything goes back to the hardware.”

“The flexibility of Apple silicon architecture remains one of their biggest technical advantages over competitors,” said Ben Bajarin, Creative Strategies analyst.

“Apple is getting a lot faster at updating its silicon with the announcement of the M4 chip. This is important as more AI features are run on device, and Apple looks to make hardware an AI competitive advantage,” wrote Deep Water Management analyst Gene Munster

“We believe the seeds for an Apple growth turnaround are being planted in the field by Cook & Co.,” Wedbush analyst Daniel Ives added. (Wedbush had previously said: “We believe the worst is in the rear-view mirror for Apple and now there is a massive iPhone product cycle with pent-up demand and AI now coming to Cupertino looking out over the next 6-9 months.”)

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