Apple: A Practical Approach to Innovation
“Apple does things practically.” Or, “Apple is late because it’s perfecting the tech.” “Would you prefer being the first or the best?” These are just a few arguments that often fuel heated discussions on Reddit or social media, especially when people are seeking a spark of outrage. Yet, there is a kernel of truth in these sentiments, as many innovations take years, if not decades, to develop. One exemplary initiative is Apple Intelligence, which is being viewed as Apple's response to the rising trend of generative AI.
Apple Intelligence made its debut with the iOS 18.1 developer beta last month, granting users their first opportunity to experience it. I dove right in.
Apple has a long history with AI. A brief historical background on the journey reveals how far we've come. Remember the Apple Newton MessagePad? Its criticized handwriting recognition feature, known internally as Rosetta, was introduced in 1995, after being conceived in 1987 by Apple engineer Steve Salomon. This technology was built on the early foundations of neural networks, supported by segmentation and a language model that utilized dictionaries and probabilistic grammar. Sound familiar? This was early AI at work—similar to the terms flooding today’s social media.
By 2014, Apple transitioned Siri to leverage machine learning techniques, such as deep neural networks, natural language processing, and convolutional neural networks, without fanfare. This evolution shows that even the most oblique Apple products have been utilizing cutting-edge AI technologies for years. Whether it’s creating memory films from your best photos or adapting to your phone usage habits to prolong battery life, machine learning has been integral to the iPhone experience for quite some time.
The introduction of Apple Intelligence appears to be a strategic move to reassure shareholders who were growing concerned amidst rising awareness of AI tools flaunted across various platforms. Thus, it’s not surprising that the announcement didn’t generate significant excitement at WWDC 2024. Apple is taking a practical approach, framing it as “AI for the rest of us.”
Apple Intelligence offers a variety of functionalities. It can summarize websites, intelligently manage notifications, generate images, and more. A key focus is improving writing skills through Apple’s new Writing Tools—a suite of features designed to help you articulate thoughts clearly, according to Apple. This has been my main focus while exploring Apple Intelligence. Here are my first impressions.
“Everywhere words matter,” declares Apple’s tagline for Writing Tools. Is it effective? It works, but I wouldn’t rely on it for professional tasks. There are a couple of key issues. First, it often misses the crux of the message. Second, it presents numerous ethical challenges. For example, when I used it to draft an email to an academic seeking expert advice, Apple Intelligence produced a grammatically flawless rewrite but entirely overlooked the two most pertinent points I wished to discuss.
This was my initial experience with Writing Tools. Unfortunately, it didn’t improve from there. After selecting the content and choosing the “Professional” preset, I expected Apple Intelligence to enhance the tone of my message. Instead, it omitted the main subject and completely altered the request. My original email was about scheduling an interview and asking for research recommendations, but the output read: “I am writing to express my interest in the opportunity to intern with your esteemed institution.” Apple’s summarizer consistently overlooks key details, especially with its Summary, Key Points, and List functionalities present in the Writing Tools suite. The shortcomings persist across both Apple’s native apps and third-party platforms like Gmail.
For instance, I tried summarizing an email regarding an auction of vintage Apple items related to Steve Jobs. While the standout item was a functional Apple-1 computer, Writing Tools failed to capture this vital information. In contrast, an email summary I received from Shortwave, an email app utilizing OpenAI’s latest GPT-4 model, accurately identified the most significant details from the dense text.
While Apple’s Writing Tools show slight improvement for longer notes, they still frequently miss crucial elements. This might be related to how Apple defines an appropriate length for a “summary” while balancing necessary details for conciseness. Another drawback lies in its lack of customization and adaptability; there are no manual adjustment options in the Writing Tools kit. Apple’s interpretation of what’s “Friendly” may miss the mark for some users.
Consider alternatives like Paragraph AI, a GPT-powered writing tool available as a mobile keyboard and browser extension. It offers slider controls to adjust the writing tone across various presets, such as informal/formal and friendly/assertive. Additionally, it can transform paragraphs into lists, text messages, emails, and full articles while providing word count adjustments. The entire system operates seamlessly.
There’s certainly room for growth. As this is still in the beta phase, some imperfections can be expected. Apple is not short on expenses—having reported around $29 billion in operating cash flow for its latest quarter—or on top industry talent. This means Apple Intelligence can anticipate significant improvements before its full-scale release.
Despite Apple Intelligence’s promise of advanced functionalities, not all features will likely appeal to mainstream users in the long run. Features like Genmojis and instant custom image creation sound exciting but may not find everyday utility. My teenage brother might enjoy them, but for him, the existing array of animated GIFs and emojis is sufficient for engaging chats without needing elaborate AI-generated images. Custom images may have occasional utility but won’t replace the quick wit essential for lively conversations.
Email summaries are certainly practical, but let’s be honest—they’re intended for serious communication. While we use email for tasks such as work correspondence, academic discussions, and finances, platforms like iMessage and WhatsApp cater to social banter. The potential risks of AI misunderstandings in such serious contexts are concerning. Enjoying intelligent notification sorting also raises uncertainties; I prefer to customize my settings per app rather than entrust AI with determining importance during focus hours.
On the surface, Writing Tools seems to hold the most practical value within the Apple Intelligence suite. However, it carries ethical concerns. Over the past few months, I’ve discussed the ethics of AI implementation with various newsroom editors, and many are skeptical about integration within editorial processes—most avoid allowing AI anywhere near story creation, from drafts to edits.
So, who is the Writing Tools bundle really for? I’m uncertain. As a reporter, I struggle to see its utility, given the ethical questions involved. In my other persona as a playful commenter on Reddit and Slack, I’ll never concede that AI can out-sass me, especially when it’s less reliable than my own lapses in memory. There likely exists a demographic leading a more carefree life who can benefit from it. However, at this stage, Apple Intelligence’s Writing Tools feel like a haphazard attempt to catch up. The silver lining is that Apple has ample time to refine the tool before its public launch, and the company has a challenging road ahead.
Ranking
