投資人對萬眾矚目的WWDC 2024最常提出的問題與我的想法 / Most Frequently Asked Questions from Investors and My Thoughts About the Highly Anticipated WWDC 2024
投資人非常關注Apple將在WWDC 2024發布的AI服務和功能。以下是過去數月中投資人們最常詢問我的問題：
- 市場如何看待WWDC 2024？
市場期待Apple能重新定義裝置端AI的使用者行為並帶動硬體換機潮。Apple股價過去數週持續上漲，顯示市場對WWDC 2024持正向態度。
2. Apple的硬體產品出貨能否在推出AI服務後重返成長趨勢？
判斷方式是AI服務是否具備「獨特且可持續的競爭優勢」。以OpenAI推出的macOS桌面端ChatGPT為例，雖較Windows桌面端領先推出，但因不獨特 (功能與網頁端幾乎相同)，優勢又不可持續 (今年稍晚就會有Windows版本)，故macOS桌面端ChatGPT對Apple硬體出貨成長並沒有幫助。
3. iOS 18的AI功能是否有助於iPhone換機需求？
我最新的供應鏈調查顯示，2024年下半年iPhone 16出貨量將同比下降約5% YoY (vs. 2H23 iPhone 15出貨)，可能意味著Apple目前認為iOS 18對換機需求幫助有限。WWDC結束後，Apple是否會因市場反饋而改變出貨預估值得關注。
4. Apple的服務業務能否受益於AI服務的推出？
裝置端AI難以對用戶收費，而雲端AI或雲端與裝置端整合的AI更有機會。相比AI帶動硬體換機需求，說服Apple既有生態內的用戶為高品質的AI服務付費更為容易。
5. Apple將推出哪些類型AI服務？
可能包括：1) Siri體驗顯著改善、2) AI強化應用程式（翻譯、搜尋、照片編輯、文書處理等）、3) 整合如ChatGPT的聊天機器人以強化網路搜尋服務、與 4) 整合應用程式內的聊天機器人/AI（使用應用時無需切換到瀏覽器便可使用AI）。
6. 與OpenAI之類的合作能否提升Apple競爭優勢？
許多投資人對這種形式的潛在合作期待很高，但還是要看合作方式與產出。以上述的macOS桌面版ChatGPT為例，這也是一種Apple與OpenAI的合作，但對Mac的出貨量毫無幫助。
重點在於合作推出的AI服務/功能，是否具備「獨特且可持續的競爭優勢」，而這就意味著與Apple合作的廠商，必須在某種形式上與Apple是獨家合作，而且Apple也能部分參與控制合作廠商大語言模型的研發進度，以確保服務品質與未來升級。否則，任何合作所推出的AI服務/功能，都很容易被競爭對手模仿，優勢也無法持續。但對類似OpenAI這樣的領導廠商，上述合作條件是很難接受的。另外Apple也必須考慮到，如果跟類似OpenAI的企業有比較深層的合作，合作開發的AI服務/功能在某些市場 (如：中國) 或有無法使用的風險。
從上述推論看，Apple與OpenAI之類的廠商合作，比較可能僅停留在「AI Wrapper」的層次 (Apple設計app的使用者介面，實際推理靠合作廠商的AI模型與算力)。以Apple的設計能力，一定能做好AI Wrapper的用戶體驗 (如設計良好的介面、便利使用等)。
7. Apple是否正在開發自己的大語言模型？
最新調查顯示，Apple正在開發自己的大語言模型，包括雲端和裝置端，但WWDC 2024可能不會有開發進度顯著優於市場預期的成果宣布。雲端大語言模型的訓練需要更多時間，而目前裝置端大語言模型的開發則受限於iPhone 16的8GB DRAM。
8. 總結Apple在WWDC 2024的AI佈局
1) WWDC結束後，Apple將不再被視為AI產業的落後者。
2) AI將在Apple主要的應用於程式中無所不在，Apple的用戶介面設計優勢將充分展現，Siri的改善是最大賣點之一。
3) 與OpenAI之類公司的合作可能主要是AI Wrapper模式。
4) 裝置端大語言模型因iPhone 16的8GB DRAM限制難超預市場預期。
5) 目前預估2024年下半年iPhone 16出貨量將低於2023年下半年的iPhone 15。
6) 長期來看，正向看待Apple對AI的投資有助於硬體換機潮與服務業務，但短期內可能不容易看到AI對營收與利潤有顯著幫助。
Investors generally focus on the many AI services and features Apple will announce at WWDC 2024. Here are the questions I have been asked frequently by investors over the past few months.
1. How does the market view WWDC 2024?
The market expects Apple to redefine user behavior with on-device AI and drive hardware replacement demand. The continuous rise in Apple’s stock price over the past few weeks indicates a positive market sentiment toward WWDC 2024.
2. Can Apple’s hardware shipments return to growth after the launch of AI services?
The key is whether the AI services have “unique and sustainable competitive advantages.” Take the macOS desktop version of ChatGPT launched by OpenAI as an example. Although it was launched before the Windows desktop version, it is not unique (the features are almost the same as the web version), and the advantage is not sustainable (a Windows version will be available later this year). Therefore, the macOS desktop version of ChatGPT does not contribute to Apple’s hardware shipment growth.
3. Will the AI features of iOS 18 help drive iPhone replacement demand?
My latest supply chain surveys indicate that iPhone 16 shipments in 2H24 will decline by about 5% YoY (vs. iPhone 15 shipments in 2H23), meaning that Apple may currently believe that iOS 18 will have limited help in driving replacement demand. It remains to be seen whether Apple will change its shipment estimates based on market feedback after WWDC.
4. Can Apple’s services business benefit from introducing AI services?
On-device- AI is difficult to charge users for, while cloud-based AI or integrated cloud and on-device AI have better opportunities. Convincing existing users within the Apple ecosystem to pay for high-quality AI services is easier than expecting AI to drive hardware replacement.
5. What types of AI services might Apple release?
Services that could be released include:
1) Significant user experience improvements to Siri.
2) AI-enhanced applications (translation, search, photo editing, word processing, etc.).
3) Integration of chatbots such as ChatGPT to enhance internet search services and
4) In-app chatbots/AI (allowing AI to be used within apps without switching to a browser).
6. Can partnerships with companies like OpenAI significantly enhance Apple’s competitive advantage?
Many investors have high expectations for such potential partnerships, but it depends on the cooperation models and outputs. For example, the macOS desktop version of ChatGPT is a collaboration between Apple and OpenAI, but it doesn’t help Mac shipments.
The key is whether the AI services/features developed through the collaboration have “unique and sustainable competitive advantages.” It means that partners working with Apple must enter into some form of exclusive collaboration and allow Apple to have some control over the partner’s Large Language Model (LLM) development progress to ensure service quality and future upgrades. Otherwise, any AI services/features resulting from the collaboration can be easily imitated by competitors, making it challenging to maintain a competitive edge. However, for leading companies such as OpenAI, such terms of cooperation are currently difficult to accept. In addition, Apple needs to consider the risk that by collaborating more closely with companies like OpenAI, the AI services/features developed through the collaboration may not be available in certain markets (such as China).
From this analysis, it seems that Apple’s collaboration with companies like OpenAI is more likely to remain at the level of “AI wrappers” (Apple designs the user interface of the apps, while the actual inference relies on the AI models and computing power of its partner companies). Given Apple’s design capabilities, they can certainly deliver an excellent user experience for AI wrappers (e.g., well-designed interfaces, ease of use, etc.).
7. Is Apple developing its own LLM?
My latest survey indicates that Apple is developing its cloud-based and on-device LLMs. However, it is unlikely that Apple will announce development progress that significantly exceeds market expectations at WWDC 2024. Cloud-based LLMs require more time to train, and the development of on-device LLMs is currently limited by the iPhone 16’s 8GB of DRAM.
8. Summary of Apple’s AI strategy at WWDC 2024:
1) After WWDC, Apple will no longer be seen as a laggard in the AI industry.
2) AI will be ubiquitous in Apple’s main applications, showcasing Apple’s advantage in user interface design. Improvements to Siri will be one of the key highlights.
3) Collaborations with companies like OpenAI may focus mainly on AI wrapper models.
4) On-device LLMs are unlikely to exceed market expectations due to the iPhone 16’s 8GB DRAM limitation and,
5) Current estimates suggest that iPhone 16 shipments in 2H24 will be lower than iPhone 15 shipments in 2H23, and,
6) In the long term, Apple’s investments in AI are seen as positively contributing to hardware replacement demand and the service business, but in the short term, significant revenue and profit growth due to AI may take more work to achieve.
