Apple shares opened the new year 2024 with a fall of more than 5%, probably the worst start to the year since 1982 or in 42 years.
What investors and Apple fans have been watching and feeling is a small knot of the past year is the AI world.
OpenAI, Copilot, Bard, and Gemini are all known to be sticking to technology headlines throughout the year.
As for Apple, there is only one hot news, namely Apple Vision Pro, besides that, the trend of iPhone 15, iPad, and MacBook is also called a slump. Not strong. It does not create excitement and has lost its magic.
As a result, Apple's long-standing market cap title began to shake for the first time.
Although Apple shares have grown by more than 50% in the last 12 months, they are considered to be lagging behind AI-related tech giants such as Nvidia, Amazon, Microsoft, and Alphabet.
But in reality, Apple is not considered to be behind the AI train as most people understand, but it may be Apple itself, which does not use the word AI, but instead calls it machine learning, and it has been used since iOS 10 (2016 or 8 years ago, long before others).
Starting from combining it with AR functions (which we tried to push a lot from the beginning until now), and let's get the most out of it in iOS 11 with CoreML inserted, developers can train their own machine learning models into their apps.
It was quite advanced at that time.
It is said that at the Earnings Call for Investors, Apple only said the word AI 10 times, compared to Amazon (50 times), Microsoft (70 times) and Google who said the word AI 90 times.
So it's not surprising that investors are questioning whether Apple has fallen behind the AI car.
But that's not the case.
Apple is out of the Generative AI trend, this is a true story. In addition to rumors that insiders are developing their own LLM model called "Ajax" or nicknamed AppleGPT, running with over 200 billion parameters and still in the experimental stage.
In addition, Apple has no news about AI, and Siri, who was a hero many years ago, shows no signs of being upgraded to be smarter and can fight ChatGPT until an internal AI Summit was held and talked about Smarter Siri.
What Apple is doing right now is called "Edge AI" or bringing embedded AI into every device it has.
Anyone who follows CES 2024 news will know that all high-tech equipment at the event has the word AI pasted on it and includes AI functions as a selling point.
Apple's Edge AI, or Device AI, is being developed to be as smart, smart, and powerful as OpenAI's GPT model
But instead of running on the cloud like everyone else, Apple can make it run on iPhone, iPad, Apple Watch.
It's starting to sound a little cooler.
Edge AI will become the core of the devices that will be released this year. Its selling point is that it is faster because it runs on our machine, not the cloud, there are no freezes, and it is slow to process.
It is more personalized because it depends on the device we use each person, and it is safer because it does not share our information or take our personal information to training. AI
But intelligence may not be as good. Because the data used is less trendy from the privacy concern that Apple pays so much attention to. To take the user's personal usage data to train.
Edge AI is being developed
The emergence of Edge AI is still in a narrow circle. People don't pay much attention to it, which may cause people to overlook Apple, but Apple itself is quietly lurking, such as:
- Apple Partners with Columbia University In October 2023, Ferret released an open-source multimodal LLM, which is an on-device LLM. No need to use the cloud, but it is very quiet news.
- LLMs have been developed to be used on devices with limited memory and stored on Flash Memory.
- Until a French startup called Mistral released an LLM model that ran on a Mac using an M2 CPU and achieved a performance similar to OpenAI's GPT3.5. There is a possibility that it will be used by companies like Apple.
- There is a lot of interest in Google to include the Gemini model in the Pixel Pro phone.
- Apple then began to release on-device AI research called HUGS, which creates 3D avatars into videos using devices like the iPhone and iPad, which is 100 times faster than traditional 3D avatar creations called NeuMan.
- LLM models are run with billions of parameters on devices like iPhones and they work fine.
Reference Research Topic " LLM in a flash ”
This is just publicly available information. It shows that Apple is quietly working on Edge AI.
Apple's control of everything
- Hardware such as mobile phones, tablets, computers, watches, headphones.
- Software for various OSs such as iOS, iPad OS, MacOS, WatchOS, tvOS.
- Chips (Apple Silicon M family, A family),
- Other chipsets (e.g., those that control photography, machine learning command processing)
- Includes services like iCloud, iWorks, Apple TV+, Apple Music, and
- SDK that developers can use right away.
It should be a point that can differentiate itself from the GenAI of competitors.
Analysts expect Apple to make a comeback in AI because the ecosystem is very strong in both hardware and software, which we should see seriously. In the launch of the iPhone 16 at the end of the year (it's been a long time, 55)
New York Times sues Microsoft and OpenAI for allegedly taking content Article: I went to train my own AI.
But Apple uses a forward game approach, talking to major publishers on its own. No risk of being sued.
Regulators in Europe, the UK, and the U.S. are turning to OpenAI to control content, data accuracy, and adoption.
Apple is watching and waiting to see how much GenAI will be controlled and interfered with by various agencies.
Another thing that Apple is doing quietly and without much shakiness is that the Apple Watch, which has become the most used health device in the world right now, will also be packed with AI functions.
Disciples like us. There is hope and enough to stretch your chest.
Let's wait and see if the strategy that Apple has always used in the Tim Cook era will work as well as intended.
Opinion of Insiderly.ai
- This year, Apple's AI should be released in the whole panel and should be as powerful as ChatGPT.
- Mass adoption can also occur quickly. If you look at the number of Apple devices available in the market, all categories combined.
- Apple's AI strategy is slow but best by focusing on its own strengths, the ecosystem, without following the games that others are competing with (whether it will succeed remains to be seen).
- There is also Apple Vision Pro that has started to go on sale and the feedback is very good. There is an opportunity to integrate AR/VR/XR into an Apple-style Metaverse, which when combined with Edge AI, should lead to applications related to the Metaverse and AI, which no one has done before.
- It's a very interesting and interesting strategy.
Source
