☕️ Apple's First Big Steps Into AI @ WWDC 2024

Apple kicks the week off with its Worldwide Developer Conference (WWDC) 2024, going over the company's most recent developments, from Apple Intelligence to potential future partnerships. Meanwhile, Microsoft finally addresses security concerns regarding Recall, Adobe slips into controversy with its latest Terms of Service changes, and Meta sets itself up to train an undisclosed AI on European users' social media posts. Join us at AI Tangle as we untangle this week's happenings in AI.

Everything announced at the Apple WWDC 2024 keynote

The long-awaited Worldwide Developer Conference 2024 is finally upon us, and with it came high expectations as to whether Apple, a company that, unlike its rivals, hasn't been at the forefront of AI development, can meet them. As usual, the event started with a keynote announcing an array of things to come, from software and hardware updates to teasers of new originals to release on the company's Apple TV+ streaming service.

​

What's the summary of the reveals?

One of the main highlights of the WWDC 2024 keynote is Apple Intelligence, the company's main AI-centered feature with safety and a high degree of personalization at its core, supported for devices with at least an M1 chip, like the iPhone Pro 15 and M1 Mac. Apple's iOS 18 would be the next centerpiece, one of the iPhone's most highly-anticipated updates in its history, which includes bringing ChatGPT to Siri, something that Elon Musk isn't too happy about as he threatens to ban Apple's devices from his businesses.

Other announcements made at the keynote are:

• AppleTV+'s InSight feature that allows users to learn actor names and song titles as they appear on the screen

• Plans to work with Google's Gemini "in the future"

• ​visionOS2, a massive update to the operating system running on the Vision Pro

• The latest macOS update, dubbed Sequioa​

• iPadOS 18 features like Smart Script and Math Notes​

• ​The Passwords app, available with the next major releases of iOS, macOS and visionOS

• A makeover to the Photos app for iOS 18

• ​Tap to Cash, which allows iPhone users to pay for things by tapping iPhones together

Microsoft revamps Recall's security mechanisms after facing heavy online backlash

After facing vocal concerns about not just privacy but also cybersecurity surrounding its Recall feature, Microsoft is making it an opt-in feature in Windows 11, which was originally on by default. The Recall feature, which comes as part of its upcoming Copilot Plus PCs, screenshots user activity for easy backtracking purposes later, and will require Windows Hello authentication for access after the changes. Despite concerns about data stored in plain text, Microsoft assures that all data remains local and private by adding more layers of data protection.

​

Adobe's latest Terms of Service changes to train its Firefly AI causes widespread controversy

Adobe's recent Terms of Service update sparked public outrage as users interpreted it as Adobe claiming rights to use their work for training the company's AI, Firefly. In a blog post​ aimed to clarify this, Adobe stated that it doesn't train its AI model on customer content, with Adobe's Chief Product Officer stating that the unclear wording in the TOS would be revised. Adobe's image in the past few years has left many, especially individual creators, distrustful of the company, and this incident adds to the pile as many point out that the released blog post doesn't even cover the main point of contention surrounding Section 4.2.

Update: Another blog post from Adobe aims to rectify the shortcomings of the previous.

​

Meta aims to train its latest AI using social media posts across Europe

Meta, the parent company of social media giant Facebook, has announced plans to use public social media content from Europe to train its AI models, a practice it already does globally. Despite initial concerns over EU privacy regulations, Meta will use public posts while excluding private ones to comply. European users will also be notified about its use of public data for AI development. However, advocacy groups like NYOB (none of your business) have challenged this move across many European countries, stating that Meta needs to obtain opt-in consent from users and that simple notifications are insufficient.

​

Nvidia set to go through a 10-to-1 stock split after a generational rally, what's next for the AI market

Nvidia will undergo a 10-for-1 stock split following a 212% increase in value over the past year, which propelled its value past $3 trillion and thus overtaking Apple as the second most valuable company in the world. Experts, like Bank of America's Vivek Arya, view Nvidia as a "generational opportunity" and anticipate further growth, particularly as the AI industry continues to expand, as the company's gains have ​single-handedly contributed more than 30% to the S&P 500's returns​ this year.

​

China's Guangdong province rolls out 45 new measures to boost China's AI industry

The Guangdong province in China has recently introduced 45 measures to boost AI use across various sectors, aiming for the AI industry's value to reach $41.4 billion by next year and $60.7 billion by 2027, also aiming to surpass 40 exaflops of computing power and 60 exaflops by 2024 and 2027 respectively. Guangdong's measures include the development of a mature AI industrial chain, fostering an innovative AI chip ecosystem, and attracting top-tier AI talent within the region.

​Riffo - Riffo is a simple AI tool with a simple purpose that renames gibberish filenames on your documents, pictures, and videos clearer and more concise ones to reduce clutter.

​Humanize - Humanize is a text transformer tool to make AI-generated content sound more human-like by combining sentiment analysis and contextual intelligence.

​Text Blaze - Automate your workflow with Text Blaze, a writing assistant that saves you work hours by helping summarize data and meeting info, create personalized messages, and schedule and automate processes with reusable text blocks and messages.

​Appinion - User appinion matters, so get your feedback from Appinions' AI models that give proactive feedback, automate qualitative data analysis, and give insight into what you need.

​How Microsoft's Awkward Position in AI Could Help Google (4-min read)

Todd McKinnon, CEO of identity security tech firm Okta, chimes in on how Microsoft's role in the AI industry is turning into a sort of "bizarre consultancy," and why he thinks it could help Google slowly achieve commercial success with its own AI investments.

US or China Brand AI & Employment in The Future (26-min watch)

Kai-Fu Lee, an AI researcher, entrepreneur, and CEO of 01.A1, shares his views on the different yet parallel universes that the US and Chinese are developing AI on, what current tech giants will remain or rise to the top in the AI market, and how employment will rapidly change in the near future.