Apple, which has been frequently searched in recent days, seems to have reached a major juncture of transformation.
On February 28, it was reported that Apple would disband the 10-year-old car project team and completely halt this long-term research and development plan. On the same day, Apple's shareholder meeting was held as scheduled, and CEO Tim Cook did not take the initiative to mention the situation of the automotive business at the meeting, but talked about generative AI technology.
One of the most notable is Cook's statement that he will invest heavily in the AI field, and revealed that he will show the outside world later this year how AIGC is breaking new ground. It is worth mentioning that when the car business was shut down, it was revealed that Apple would transfer relevant resources to generative AI.
In 2023, when AI models are running blindfolded, Apple's attitude towards this technology can be described as ambiguous. Cook himself has rarely spoken publicly about the company's investments and research and development plans in the field. Now it is unusual, indicating that Apple has finally begun to withstand the impact of the giant wheel of the times and has begun to face a new round of technological revolution in the technology circle.
It's just that Google, Microsoft, Meta and other Silicon Valley counterparts have been working hard for a whole year, can Apple, which is long overdue, really catch up?
*From pixabay).
Objectively speaking, Cook has no shortage of confidence. Although Apple has not invested heavily in generative AI before, it has a very extensive layout in the field of AI, and has accumulated a large number of technologies and products through decades of research and development.
Apple's earliest and most popular product in the field of AI is undoubtedly Siri. Although it was acquired from outside, Siri's current status is due to Apple's transformation. It is under the leadership of Steve Jobs that Siri's positioning has changed from a simple voice assistant to an intelligent interaction method, breaking free from the functional limitations of search.
Siri's technological leap forward came with the launch of iOS 10 in 2016. This year coincided with Google's AlphaGo sweeping the Go world, bringing unprecedented popularity to AI, and Apple's response speed can be described as superb. iOS 10 also has a lot of innovations, including extending SiriKit to third-party applications, supporting photo album face recognition, etc., to penetrate AI technology into more fields. Siri's biggest improvement has been in the ability to reason, understand, and be open.
On the one hand, the third-party APP is fully accessed, and users can perform services such as payment, taxi, and voice call by waking Siri, which greatly enhances the practicability. Siri, on the other hand, can automatically associate the next step after following the instructions, such as automatically calculating the travel time and estimating the cost after entering the taxi command and destination. It can be said that Siri at this time is already very close to the positioning of intelligent assistants.
In 2018, Apple's AI landscape opened a new chapter: Core ML and Create ML were launched, which means that Apple's AI development ecosystem is open to the outside world.
At this year's WWDC Worldwide Developers Conference, machine learning and artificial intelligence also became the absolute protagonists. In addition to these two open-source tools, Apple has also launched hardware products such as smart speakers HomePod, and revealed that it will accelerate the research and development process of AI chips. To this end, Apple has also hired Google's head of search and artificial intelligence to be responsible for machine learning-related projects.
The ML framework, the exclusive machine learning framework for the M series chips released last year, is Apple's core in the era of large models.
According to official data, the ML architecture can be directly used to train 7 billion parameter large models, including the most common Transformer architecture language models. The performance of this architecture has been praised by a large number of industry leaders, including Turing Award winner Yann Lecun and NVIDIA senior scientist Jim Fan, who has called Apple's biggest breakthrough in the field of AI open source.
What's more, with the ML architecture, the efficiency will be greatly improved, whether it is the optimization of AI chips or the development of generative AI applications for Mac.
Finally, I have to mention the Apple car project, which is now "uncertain". You know, Cook once called the project the "mother of AI", because it brings together Apple's best engineering and algorithm perception teams, and the hardware and systems required for smart cars are inseparable from AI.
According to public information, although the Apple car team has a frequent turnover of personnel, it has been recruiting R&D talents for three-electric technology and autonomous driving systems since 2015. Among them, there are a large number of algorithm and software engineers poached from Tesla, Samsung, Texas Instruments and other large manufacturers. Although the Apple car is likely to be the first to fight, it still leaves a lot of valuable wealth for Apple, and it can also provide a lot of help for the generative AI attack journey.
In general, Apple does not lack AI talents, technical precipitation, and money. Now that Cook has stated that he wants to increase investment, I believe that he will soon see the next move.
So here comes the new question: Next, what is Apple going to do with generative AI?
At Wednesday's shareholder meeting, Cook was full of praise for generative AI, but when asked about how to apply AI technology to specific businesses and the ethical guidelines for using AI, Cook could not give a very detailed and precise response.
Cook focused on old topics such as the Vision Pro's hand tracking tools, the Apple Watch's heart rate alert feature, and the Mac's AI chip, which are just one part of Apple's vast business landscape. Among them, the Mac is Cook's most satisfying product, which he calls "the best AI PC on the market".
Although Cook did not say it explicitly, the new trend of AI + hardware must be an important reason why Apple is actively engaged in the construction of generative AI.
From last year's CES to this year's MWC, AI hardware has become presence-ready, including AI mobile phones, AI PCs, AI smart wearable devices, and more. These events have always been the vane of the technology circle, and this scene also shows that AI+ hardware has become a battleground for soldiers.
According to Lei Technology, in MWC 2024, which is being carried out in Spain, Samsung brought the much-talked-about Galaxy Ring smart ring and announced the new positioning of the "AI ring". It is reported that in addition to the built-in ECG sensing, blood flow sensing and other functions, the device will provide users with basic services such as health monitoring, stress testing and nutrition observation, and will also add more AI functions, but the specific details have not yet been announced.
At the same time, Lenovo also brought a ThinkBook transparent notebook AI PC with a sense of technology to this MWC. Prior to this, Lenovo has released a series of AI PC products such as the ThinkPad T14 T14S 5th Gen, ThinkPad T16 3rd Gen, ThinkPad X12 Detachable 2nd Gen, and ThinkBook 14 2-in-1***.
The importance of hardware to Apple is self-evident, and the wave of AI is certainly not to be missed.
Secondly, the importance of AI to the software and application ecosystem is also an issue that Apple must consider.
Microsoft, Google, and Meta were the first to integrate AI model technology into front-end businesses such as search and social networking, and penetrated into the field of cloud services on the B-side. The increasing popularity of large models not only gives Microsoft the courage to challenge Google's search dominance, but also makes the market value of these companies all the way, which shows the importance of AI in software and applications.
Apple's ambitions were actually revealed very early. After the sudden new product launch conference in October last year to update the Mac product line, it was revealed that Apple will use the next generation of Mac as a testing ground for AI software and systems. The reason why I chose Mac is mainly because of the computing power of the chip. Apple's M3 chip has the highest computing speed and operational efficiency among its competitors, allowing it to handle AI tasks more stably.
In addition, iOS 18 and iPadOS 18, which deeply integrate AI large models, have long been put on the agenda. Haitong analyst Jeff Pu pointed out in a research report updated at the end of last year that Apple has built hundreds of AI servers in 2023 to improve edge AI processing power, and will continue to expand this year. The improvement of cloud computing power is likely to be a preparation for embedding self-developed large models in iOS 18.
All in all, the future software and hardware industry is inseparable from the keyword AI, and it will inevitably continue to penetrate Apple's various businesses.
However, Apple's AI race will not be easy.
Many would argue that falling behind the other giants in Silicon Valley is the first hurdle Apple faces on the generative AI track.
However, this statement is scored on both sides.
On the one hand, how to judge backwardness, the criteria vary from person to person. It is true that Microsoft, Google, and Meta are all more powerful than Apple, launching AI model applications such as chatbots, and integrating technology into various businesses. But Apple just doesn't deliberately mention AI, not ignoring AI.
On the contrary, Apple's products are indispensable to AI, such as the aforementioned Siri and Mac, as well as the motion tracking, object detection and other technologies applied in Vision Pro, but they did not rely on AI when marketing momentum.
On the other hand, even if it is temporarily behind, Apple may not be without the strength to fight. As mentioned above, Apple has no shortage of technology, talent, or money, and has all the necessary foundations to develop generative AI technology at scale.
In fact, what Cook should really worry about is not technology and resources, but the mistakes Apple has made with projects like cars and XRThe vacillation of the strategy and the serious closure of the ecology are the main reasons why Apple missed the first wave of new energy vehicles and XR.
Taking the car-making plan as an example, Apple has experienced several strategic transformations, each of which has delayed the R&D process to varying degrees, lost the best manufacturing personnel, and even deepened internal contradictions.
In January 2016, Steve Zadesky, Apple's vice president of product design, who was in charge of the original car-building project, announced his resignation due to the team's inability to agree on a product plan, and his position was not replaced by Bob Mansfield, a former executive who returned to Apple until six months later. Under the latter's leadership, Apple finally decided on a strategy to develop towards fully autonomous driving, and began to build a new team and apply for testing. However, at this time, Apple will focus on the research and development of autonomous driving solutions, and has not increased vehicle manufacturing.
Unexpectedly, just over a year later, Apple's strategy changed again. The car-making team reduced nearly 200 employees through job transfers and layoffs in half a year, entering a stage of comprehensive contraction, claiming to concentrate on a few core technology areas. What's even more embarrassing is that Apple actually brought back Doong Field, who had served as vice president of hardware engineering and was in charge of Model 3 production at Tesla, at about the same time, to rekindle the enthusiasm for car building.
After that, I believe everyone knows very well, it is nothing more than a repetition of the plot again and again. What Apple's internal attitude and positioning towards AI was before, the outside world has no way of knowing. But judging from the results of missing out on the first wave of large model dividends last year, Apple's strategy is certainly not too successful.
Interestingly, Cook repeatedly emphasized at the shareholder meeting that Apple has been working on "AI-driven" technology for many years and has avoided using the term "AI" alone. According to Cook, Apple has a number of products that have redefined the concept of "AI-driven" and have not abandoned Apple's long-sought "machine learning" theory. Cook's move can be said to highlight Apple's differentiation, but it is not a kind of stubbornness.
To put it mildly, Apple, which once subverted Silicon Valley with disruptive ideas and forward-looking technologies, is now living more and more like a closed, old-fashioned old aristocrat, defending its own set of unfashionable rules and ideas, trying to fight against the world. But historical experience has shown time and again that most of this seemingly fearless struggle is a futile struggle.
Fortunately, Apple has gradually made changes. From increasing open source efforts to Cook taking the initiative to mention AI at the shareholder meeting and promising to increase AI investment, Apple is lowering its noble head and learning to integrate into the trend of the world.
If you want to go beyond the trend and reach the future in advance, you must first keep up with the trend. Willing to make a change, Apple is a big step forward.