Newsletter: Why Apple is Quietly Buying AI Companies

Why Apple is Quietly Buying AI Companies

In a fictional scenario presented in a Cold Fusion video, the protagonist uses Siri to plan a date night with his wife. While this scenario is not currently possible, recent developments in generative AI suggest that it could become a reality. The video highlights the potential of voice-first AI-powered devices like the Rabbit R1, which can interface with apps to complete tasks on behalf of users.

Apple, a company known for its silence on AI, has actually bought the most AI startups since 2017, spending around $1 billion in AI Ventures. The company has been researching generative AI for years and has assembled a foundational models team to develop conversational AI with a focus on on-device machine learning and privacy. The team includes former Google engineers and the head of the team, John Jan Andrea, was previously head of machine intelligence research and search at Google.

Apple has quietly released frameworks and model libraries designed to run on its chipsets and may even bring generative AI apps to MacBooks. The company has also created a large language model called Ajax GPT, which is said to outperform GPT 3.5. There are even rumors that Apple is working on an AI-powered chatbot dubbed Apple GPT, based on Ajax.

Despite these developments, Siri, which was acquired by Apple in 2010, has been frustrating users for years. The Siri team within Apple has devolved into a dumpster fire, with employees and management bitterly arguing about the direction of the assistant. Engineers were not given tools to see how many users were using Siri and how often, and executives dismissed back-and-forth responses to Siri because they wanted Siri to appear “all-knowing.”

However, there are signs that Apple is trying to turn things around. The company has been working on rebuilding Siri from the ground up with projects like Blackbird and Siri X, which focus on making Siri faster, lighter, and more modular. Apple executives have also been reportedly pleading with disgruntled engineers to stay at the company and work on AI.

Despite the challenges, Apple’s efforts in AI are evident, and the company’s silence on the matter suggests that it is taking a measured approach to its AI development, as it has done with other technologies in the past. As the technology continues to evolve, it remains to be seen how Apple will integrate AI into its products and services in the future.

The Rabbit R1 and Its Limitations

The Rabbit R1 is a voice-first AI-powered device that aims to interface with various apps to complete tasks on behalf of its users. However, it has a fatal flaw that limits its functionality. All of its features could easily be integrated into a smartphone, rendering the Rabbit R1 unnecessary.

Although generative AI breakthroughs have made it technically possible for devices like the Rabbit R1 to execute a series of tasks automatically, it is not yet a reality. Companies like Google and Samsung have already integrated AI into their smartphones, and Apple has been silently working on its own AI efforts behind the scenes.

Apple has been buying AI startups since 2017, spending around $1 billion in AI ventures. Its research on generative AI has been ongoing for years, and it has assembled a team of 16 people, including former Google engineers, to develop conversational AI similar to that of Chat GPT. The team is focusing on on-device machine learning and privacy.

Apple has created a large language model called Ajax GPT, which is similar to Chat GPT but is said to outperform GPT 3.5. There are also rumors about Apple working on an AI-powered chatbot dubbed Apple GPT, based on Ajax. However, it is still tightly controlled within the company, and there is no confirmed release date yet.

Apple’s AI efforts have been focused on various specialties, including self-driving technology, voice design, music generation, and image recognition. The company has been quiet about its AI efforts, but it moves last and decisively. Although Siri is currently next to useless, Apple is working on improving it, and there are multiple teams working on various AI projects within the company.

The Siri team within Apple has had a lot of problems over the past five years. Employees and management would bitterly argue about the direction of the assistant, and in the end, no meaningful progress was made. Siri engineers weren’t given tools to see how many users were using Siri and how often they were using it. Apple executives wanted to keep future developments of Siri on device, putting Apple’s efforts at a disadvantage. Apple executives dismissed back and forth responses to Siri because it would be too difficult to control.

Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated. Instead, they opted for a team of 20 writers to come up with canned pre-written responses. Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses.

There was even an attempt to rebuild Siri from the ground up with a project called Blackbird. The project aimed to make Siri faster, lighter, and more modular. However, it was axed by two senior leaders on the Siri team who championed their own project, Siri X. This only focused on moving Siri’s processing to on-device, and the rest of the Blackbird project was abandoned.

Apple’s mixed reality headset team, The Vision Pro, were reportedly so unimpressed with Siri’s capabilities that they considered creating their own voice system. Despite all these issues, Apple is still working on improving Siri, and with the company’s history of revolutionizing categories, it’s possible that they will do it better than before.

Smartphone Integration Prospects

Apple has been silently working on integrating generative AI into their products, including their voice assistant, Siri. The company has been buying up AI startups since 2017, spending around $1 billion in AI Ventures. Meanwhile, Google and Microsoft have been more vocal about their AI efforts.

In August of 2023, Tim Cook revealed that Apple had been researching generative AI for years. The company has assembled a foundational models team, consisting of 16 people, including former Google engineers and the head of the team, John Jan Andrea, who was previously head of machine intelligence research and search at Google. Bloomberg reported in mid-2023 that Apple has created a large language model called Ajax GPT, which is said to outperform GPT 3.5.

There are even rumors about Apple working on an AI-powered chatbot dubbed Apple GPT, based on Ajax. Although there is a rumored release date of mid-2024, it is still tightly controlled within the company. Apple has also quietly released frameworks and model libraries designed to run on its chipsets and may even bring generative AI apps to MacBooks, according to The Information.

Multiple teams are working on various AI projects within Apple, including Siri. Siri itself was an Apple acquisition back in 2010, but it has been frustrating users ever since. The Siri team within Apple has devolved into a dumpster fire, with employees and management bitterly arguing about the direction of the assistant. In April of 2023, a bombshell report came to light, explaining all the problems within the Siri team.

Siri engineers weren’t given tools to see how many users were using Siri and how often they were using it. Apple executives wanted to keep future developments of Siri on device, having no network connection to a massive AI model run on a server put Apple’s efforts at a disadvantage. Apple executives dismissed back-and-forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated. Instead, they opted for a team of 20 writers to come up with canned pre-written responses.

Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses. In 2019, Apple’s design team repeatedly rejected a feature where users could report a concern or issue with a Siri answer. The machine learning engineers were essentially working blind and couldn’t improve the system.

Despite all these issues, Apple is working to turn things around for Siri. The company has been researching generative AI for years and has created a large language model called Ajax GPT. Multiple teams are working on various AI projects within Apple, and it is quite evident that there’s been a lot of activity from Apple on AI behind the scenes.

Apple’s Secret AI Developments

Apple has been quietly working on AI developments behind the scenes, despite barely mentioning AI in any of their product launches. In fact, since 2017, Apple has bought the most AI companies of all, spending approximately $1 billion in AI ventures. Apple has been focusing on specialties such as self-driving technology, voice design, music generation, and image recognition, among others.

In August of 2023, Tim Cook revealed that Apple had been researching generative AI for years. The company assembled a foundational models team in 2020 to develop conversational AI similar to chat GPT, but with a focus on on-device machine learning and privacy. The team consists of 16 people, including former Google engineers, and the head of the team, John Jan Andrea, was previously head of machine intelligence research and search at Google.

Bloomberg reported in mid-2023 that Apple has created a large language model called Ajax GPT, which is similar to chat GPT or Google’s Lambda but is said to outperform GPT 3.5. There’s even chatter about Apple working on an AI-powered chatbot dubbed Apple GPT that’s based on Ajax, although it’s still tightly controlled within the company.

Apple has also quietly released frameworks and model libraries designed to run on its chipsets and may even bring generative AI apps to MacBooks, according to The Information. Multiple teams are working on various AI projects within Apple, so it’s quite evident that there’s been a lot of activity from Apple on AI behind the scenes.

Although Siri is next to useless today, Apple is secretly cooking something up in the background to improve it. In April of 2023, a report came to light that explained everything. Over the past five years or so, the Siri team within Apple has devolved into a dumpster fire. Employees and management would bitterly argue about the direction of the assistant, and in the end, no meaningful progress was made.

Apple executives wanted to keep future developments of Siri on-device, having no network connection to a massive AI model run on a server, putting Apple’s efforts at a disadvantage. Apple executives dismissed back-and-forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated. Instead, they opted for a team of 20 writers to come up with canned pre-written responses.

Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses. In 2019, Apple’s design team repeatedly rejected a feature where users could report a concern or issue with a Siri answer. The machine learning engineers were essentially working blind and couldn’t improve the system.

Despite all these issues, Apple can turn things around for Siri. The company has been working on generative AI, and multiple teams are working on various AI projects within Apple. While Siri’s current state is not ideal, the company’s efforts in the AI space could lead to significant improvements in the future.

Comparative AI Company Acquisitions

Apple, Microsoft, Meta, and Google SL Alphabet are among the big tech companies investing in artificial intelligence (AI) startups. Since 2017, Apple has acquired 21 AI companies, while Microsoft has acquired 12, Meta has acquired 11, and Google SL Alphabet has acquired eight. Apple has spent around $1 billion in AI ventures, purchasing startups specializing in self-driving technology, voice design, music generation, and image recognition, among others.

Apple’s acquisition strategy is typical of the company’s behavior. It moves last but decisively, revolutionizing categories by doing them better than before. While Microsoft and Google have been vocal about their AI efforts, Apple has remained quiet. However, it has been secretly working on generative AI for years. In August 2023, Tim Cook revealed that the company had been researching generative AI for years, with a focus on device machine learning and privacy.

Apple has assembled a foundational models team consisting of 16 people, including former Google engineers and the head of the team, John Jan Andrea, who was previously head of machine intelligence research and search at Google. Bloomberg reported in mid-2023 that Apple has created a large language model called Ajax GPT, which outperforms GPT 3.5. There is even chatter about Apple working on an AI-powered chatbot dubbed Apple GPT, based on Ajax. Although there is a rumored release date of mid-2024, it is still tightly controlled within the company.

Apple has also quietly released frameworks and model libraries designed to run on its chipsets and may even bring generative AI apps to MacBooks, according to The Information. Multiple teams are working on various AI projects within Apple, indicating a lot of activity on AI behind the scenes.

Despite its efforts, Siri, Apple’s voice assistant, has been frustrating users for years. A bombshell report in April 2023 revealed that the Siri team within Apple had devolved into a dumpster fire, with employees and management bitterly arguing about the direction of the assistant. No meaningful progress was made, and Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses.

Apple executives wanted to keep future developments of Siri on-device, having no network connection to a massive AI model run on a server, putting Apple’s efforts at a disadvantage. They dismissed back-and-forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated, opting for a team of 20 writers to come up with canned pre-written responses.

Apple’s mixed reality headset team reportedly considered creating their own voice system because they were unimpressed with Siri’s capabilities. Despite all these issues, Apple is still trying to turn things around for Siri.

Apple’s Generative AI Research

Apple has been quietly working on its AI efforts in the background, with a focus on generative AI. Since 2017, Apple has acquired 21 AI startups, spending around $1 billion in AI ventures. Apple has been expanding its AI expertise in various areas, including self-driving technology, voice design, music generation, and image recognition.

While Apple has been relatively quiet about its AI efforts, it has been researching generative AI for years. In August 2023, Tim Cook revealed that Apple had been working on conversational AI similar to that of chat GPT, but with a focus on on-device machine learning and privacy. The foundational models team, consisting of 16 people, including former Google engineers and the head of the team John Jan Andrea, has been working on developing conversational AI.

Apple has created a large language model called Ajax GPT, similar to chat GPT or Google’s Lambda, which is said to outperform GPT 3.5. There are even rumors about Apple working on an AI-powered chatbot dubbed Apple GPT that’s based on Ajax. Although there’s a rumored release date of mid-2024, it’s still tightly controlled within the company.

Multiple teams are working on various AI projects within Apple, and the company has quietly released frameworks and model libraries designed to run on its chipsets. Apple may even bring generative AI apps to MacBooks, according to The Information.

Apple’s AI efforts will undoubtedly improve Siri, which has been a source of frustration for users. The Siri team within Apple has had a lot of problems over the past 5 years, with employees and management bitterly arguing about the direction of the assistant. Siri engineers weren’t given tools to see how many users were using Siri and how often they were using it. Apple executives wanted to keep future developments of Siri on-device, putting Apple’s efforts at a disadvantage.

Apple executives dismissed back-and-forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated, instead opting for a team of 20 writers to come up with canned pre-written responses. Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses.

There have been attempts to rebuild Siri from the ground up, with a project called Blackbird aiming to make Siri faster, lighter, and more modular. However, it was axed by two senior leaders on the Siri team who championed their own project, Siri X, which only focused on moving Siri’s processing to on-device, abandoning the rest of the Blackbird project.

Apple’s mixed reality headset team, The Vision Pro, reportedly considered creating their own voice system because they were so unimpressed with Siri’s capabilities. Despite all these issues, Apple is still working to improve Siri and turn things around.

Apple’s Large Language Model: Ajax GPT

Apple has been quietly working on its AI efforts behind the scenes, and it seems that they have been researching generative AI for years. In August 2023, Tim Cook revealed that the company had been working on conversational AI similar to that of chat GPT, but with a focus on on-device machine learning and privacy. The company assembled a foundational models team in 2020, consisting of 16 people, including former Google engineers and the head of the team, John Jan Andrea, who was previously head of machine intelligence research and search at Google.

According to Bloomberg, Apple has created a large language model called Ajax GPT, which is similar to chat GPT or Google’s Lambda but is said to outperform GPT 3.5. There is even chatter about Apple working on an AI-powered chatbot dubbed Apple GPT that’s based on Ajax. Although there’s a rumored release date of mid-2024, right now, it’s still tightly controlled within the company.

Apple has also quietly released frameworks and model libraries designed to run on its chipsets and may even bring generative AI apps to MacBooks, according to The Information. Multiple teams are working on various AI projects within Apple, so it’s quite evident that there’s been a lot of activity from Apple on AI behind the scenes.

It remains to be seen what the best use of Apple’s AI efforts will be, but Siri is an obvious candidate. Siri, which was an Apple acquisition back in 2010, has been frustrating users ever since its incorporation into the iPhone. In April 2023, a report came to light that explained all the problems within the Siri team. Siri engineers weren’t given tools to see how many users were using Siri and how often they were using it. Apple executives wanted to keep future developments of Siri on-device, having no network connection to a massive AI model run on a server put Apple’s efforts at a disadvantage. Apple executives dismissed back-and-forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated, instead opting for a team of 20 writers to come up with canned pre-written responses.

Apple’s obsession with strict control over the user experience is at direct odds with the freedoms of generative AI responses. In 2019, Apple’s design team repeatedly rejected a feature where users could report a concern or issue with a Siri answer. The machine learning engineers were essentially working blind and couldn’t improve the system. Upper management wanted Siri to appear “all-knowing.” Frustrations began building, and employees began to leave Apple because they were too slow to make decisions about AI and were just scared of making mistakes.

Apple can turn things around for Siri by addressing these issues. The company needs to give Siri engineers the tools they need to see how many users are using Siri and how often they are using it. Apple should also consider allowing future developments of Siri to be connected to a massive AI model run on a server, which would put Apple’s efforts on a level playing field with competitors. Finally, Apple needs to embrace the freedoms of generative AI responses and allow machine learning engineers to improve the system by having users report concerns or issues with Siri answers.

The Siri Dilemma

Apple’s Siri, the voice assistant on iPhones, has been a source of frustration for many users. Despite being one of the first voice assistants on the market, Siri has fallen behind competitors like Google Assistant and Amazon’s Alexa in terms of functionality and accuracy.

In recent years, there has been a lot of activity from Apple behind the scenes in the field of AI, including the acquisition of 21 AI startups since 2017 and the formation of a foundational models team focused on conversational AI. However, Siri has not seen the same level of improvement as other AI-powered products.

A report in April 2023 revealed that the Siri team within Apple was in disarray, with bitter arguments between employees and management and no meaningful progress being made. The team lacked tools to see how many users were using Siri and how often, and Apple executives dismissed the idea of having responses generated by AI in favor of a team of writers.

Despite these setbacks, there is hope for the future of Siri. Apple has quietly released frameworks and model libraries designed to run on its chipsets, and multiple teams are working on various AI projects within the company.

To turn things around for Siri, Apple needs to address the issues within the Siri team and embrace the potential of generative AI. With the right investment and focus, Siri could once again become a leader in the voice assistant market.

Related content:

    Internal Conflicts and Siri’s Stagnation

    Apple’s Siri, the voice assistant that was once ahead of the competition, has fallen behind in recent years. In fact, a bombshell report in April 2023 revealed that the Siri team within Apple had devolved into a dumpster fire. Employees and management bitterly argued about the direction of the assistant, and no meaningful progress was made.

    One of the main issues within the Siri team was the lack of data available to engineers. They weren’t given tools to see how many users were using Siri and how often they were using it. Engineers would beg for more data to improve Siri, but leadership wouldn’t hear any of it. Apple executives wanted to keep future developments of Siri on device, having no network connection to a massive AI model run on a server put Apple’s efforts at a disadvantage.

    Furthermore, Apple executives dismissed back and forth responses to Siri because it would be too difficult to control. Tim Cook himself, as well as other senior executives, refused to have their responses AI-generated. Instead, they opted for a team of 20 writers to come up with canned pre-written responses, which explains a lot about Apple’s obsession with strict control over the user experience being at direct odds with the freedoms of generative AI responses.

    In 2019, Apple’s design team repeatedly rejected a feature where users could report a concern or issue with a Siri answer. The machine learning engineers were essentially working blind and couldn’t improve the system. Upper management wanted Siri to appear “all-knowing,” but this led to frustration building up, and employees began to leave Apple because they were too slow to make decisions about AI and were just scared of making mistakes.

    There was even an attempt to rebuild Siri from the ground up with a project called Blackbird, which aimed to make Siri faster, lighter, and more modular. However, it was axed by two senior leaders on the Siri team who championed their own project, Siri X. This only focused on moving Siri’s processing to on-device, and the rest of the Blackbird project was abandoned.

    To add insult to injury, the group working on Apple’s mixed reality headset, The Vision Pro, were reportedly so unimpressed with Siri’s capabilities that they considered creating their own voice system. All of these issues explain why Siri isn’t doing too well and is only getting more dull. One of its most common phrases today is “sorry, I don’t understand” or “here are some results I found on the web.”

    Despite all these issues, Apple has been secretly cooking up something in the background. Since 2017, Apple has bought 21 AI startups, spending about $1 billion in AI ventures. They’ve been researching generative AI for years, and in August of 2023, Tim Cook revealed that Apple had been tinkering with similar concepts way before its release. The company has assembled a foundational models team, consisting of 16 people, including former Google engineers and the head of the team, John Jan Andrea, was previously head of machine intelligence research and search at Google.

    Bloomberg reported in mid-2023 that Apple has created a large language model called Ajax GPT, which is said to outperform GPT 3.5. There’s even chatter about Apple working on an AI-powered chatbot dubbed Apple GPT that’s based on Ajax. Although there’s a rumored release date of mid-2024, right now, it’s still tightly controlled within the company.

    Multiple teams are working on various AI projects within Apple, so it’s quite evident that there’s been a lot of activity from Apple on AI behind the scenes. It remains to be seen what Apple’s next move will be to turn around Siri’s stagnation, but with their track record of revolutionizing categories by doing it better than was done before, it’s safe to say that they will make a comeback.

    Future Directions for Siri

    Despite its current limitations, Apple has been actively investing in AI technology and acquiring AI startups since 2017, with a focus on generative AI and privacy. In August 2023, Tim Cook revealed that Apple had been researching generative AI for years, with a team consisting of former Google engineers and the head of the team previously heading machine intelligence research and search at Google.

    One of the most promising developments is the creation of a large language model called Ajax GPT, which is said to outperform GPT 3.5. There is even talk of Apple working on an AI-powered chatbot dubbed Apple GPT, which is based on Ajax. While there is no official release date yet, rumors suggest it may be released in mid-2024.

    Apple has also quietly released frameworks and model libraries designed to run on its chipsets, and may even bring generative AI apps to MacBooks, according to The Information. With multiple teams working on various AI projects within Apple, it is evident that there has been a lot of activity from Apple on AI behind the scenes.

    As for Siri, the assistant has been plagued by internal disputes and a lack of data to improve its functionality. However, with Apple’s increased investment in AI technology and its focus on generative AI, there is hope that Siri will improve in the future. Apple’s strict control over the user experience may be at odds with the freedoms of generative AI responses, but with the right approach, Apple may be able to revolutionize the voice assistant market once again.

    Date and Time Display