Apple is working on a new family of multimodal models called MM1
MM1 models include as many as 30 billion parameters
The company has developed an LLM model that can interpret both images and text data, which it claims to be capable of in-context learning
Apple is working on a new family of multimodal models called MM1 that can interpret and generate different types of data at the same time. The company has developed an LLM model that can interpret both images and text data, which it claims to be capable of in-context learning, meaning it does not need to start over every time a question is asked; it uses what it has learned in the current conversation. Apple's MM1 models include as many as 30 billion parameters.
Apple is in talks with Google about licensing its Gemini technology to power some AI features coming to the iPhone
The New York Times has echoed a report from Bloomberg that Apple is discussing using Google's generative artificial intelligence model called Gemini for its next iPhone
Bloomberg reported that Apple is preparing a wide array of new AI features for iOS 18, which is set to debut at WWDC in June
Accuracy
No Contradictions at Time
Of
Publication
Deception
(100%)
None Found At Time Of
Publication
Fallacies
(70%)
None Found At Time Of
Publication
Bias
(75%)
The article reports on a partnership between Apple and Google to license Gemini technology for AI features in the iPhone. The author cites sources who claim that discussions are preliminary and have not been defined yet. However, they also mention that Apple has held talks with other AIs companies as well.
Apple is in discussions with Google about using the search giant's generative artificial intelligence model called Gemini for its next iPhone
Citing two people familiar with its development, today's report also says that Apple's effort to develop its own large language model has been running behind the likes of ChatGPT and Gemini
The exact scope of a potential deal hasn’t been defined yet
Site
Conflicts
Of
Interest (50%)
None Found At Time Of
Publication
Author
Conflicts
Of
Interest (50%)
The author has multiple conflicts of interest on the topics provided. The article mentions Bloomberg Mark Gurman and two people familiar with its development which could indicate a personal or professional relationship between the author and these individuals. Additionally, there is no disclosure of any financial ties between Apple and Google in regards to their AI partnership talks.
Apple is in talks with Google about licensing its Gemini technology to power some AI features coming to the iPhone
The New York Times has echoed a report from Bloomberg that Apple is discussing using Google's generative artificial intelligence model called Gemini for its next iPhone
To match peers like Alphabet's Google and Microsoft, Apple needs to invest heavily in AI software and data center infrastructure
Accuracy
No Contradictions at Time
Of
Publication
Deception
(50%)
The article is deceptive in several ways. Firstly, the author claims that Apple has developed an LLM model capable of interpreting both images and text data when no such model exists yet. Secondly, the author uses quotes from a paper to claim that Apple's MM1 models can count objects and identify them as part of an image without providing any evidence or context for these claims.
The article states 'Apple has developed an LLM model capable of interpreting both images and text data.' However, no such model exists yet. This is a lie by omission.
Fallacies
(85%)
The article contains several fallacies. Firstly, the author uses an appeal to authority by stating that Apple has developed a next-generation LLM without providing any evidence or context for this claim. Secondly, the author makes use of inflammatory rhetoric when they describe multimodal AI as having more comprehensive understanding and leading to more accurate interpretations than single-mode AI systems. This is not necessarily true and could be seen as a form of hyperbole. Thirdly, the author uses an example that involves uploading an image of friends at a bar holding menus which may not accurately represent the capabilities of the LLM in real life scenarios.
Apple has developed a next-generation LLM
Multimodal AI is more comprehensive and accurate than single-mode AI systems
The MMLL can count objects, identify objects that are part of an image, and use common sense about everyday objects to offer users useful information.
Bias
(85%)
The article contains examples of religious bias. The author uses the phrase 'the company is not interested in simply adding an LLM developed by another company' which implies that Apple has a preference for developing its own technology rather than relying on other companies. This could be seen as a form of religious bias, where Apple believes it should develop and control all aspects of its products.
The author uses the phrase 'the company is not interested in simply adding an LLM developed by another company' which implies that Apple has a preference for developing its own technology rather than relying on other companies. This could be seen as a form of religious bias, where Apple believes it should develop and control all aspects of its products.
Apple's recent push to catch up in the field of artificial intelligence will likely pressure its profit margins
`My guess is over the next couple years, starting in the back half of this year, we will see net income slowly decline'' - with a margin closer to 24%
Accuracy
Apple's margins have defied the laws of gravity for a decade
Deception
(30%)
The article is deceptive in several ways. Firstly, the author claims that Apple's AI initiative will be a costly undertaking and profits could take a near-term hit. However, this statement contradicts the fact that Apple has consistently maintained high profit margins over the past decade despite revenue declines. Secondly, Munster states that to match peers like Alphabet's Google and Microsoft, Apple needs to invest heavily in AI software and data center infrastructure. This is not entirely accurate as other companies have also invested heavily in these areas without experiencing a significant decline in profit margins. Thirdly, the article reports on rumors of Apple licensing Google's Gemini AI engine and discussing deals with OpenAI, which may or may not happen. The author presents this information as fact when it is merely speculation.
The company's margins have defied the laws of gravity for a decade,
Fallacies
(70%)
The article contains several examples of informal fallacies. The author uses an appeal to authority by citing the opinions of Gene Munster and other sources without providing any evidence or reasoning for their claims. Additionally, the author makes a false dilemma by stating that Apple needs to invest heavily in AI software and data center infrastructure or risk falling behind its peers, when there may be alternative options available. The article also contains inflammatory rhetoric with phrases such as 'Apple's margins have defied the laws of gravity for a decade' and 'Despite total revenue being down 2.8% in fiscal '23, net income margins were 25.3%, essentially in line with fiscal '19 record 25.9% margin.' These statements are exaggerated and do not accurately reflect the situation.
The author uses an appeal to authority by citing the opinions of Gene Munster without providing any evidence or reasoning for their claims.
Bias
(80%)
The article discusses the potential impact of Apple's AI focus on its profit margins. The author cites an investment firm that predicts a near-term hit to profits due to the costly undertaking of developing AI software and data center infrastructure. Additionally, the author mentions discussions between Apple and other companies about licensing foundation models for generative AI development.
My guess is over the next couple years, starting in the back half of this year, we will see net income slowly decline,
Apple has developed a family of multimodal models called MM1 that can interpret and generate different types of data at the same time
`MM1` offers advanced reasoning and in-context learning to respond to text and images
`MM1a could benefit future Apple products, including iPhones, Macs and Siri voice assistant
Accuracy
Apple is expected to unveil several new AI features at its developer conference in June
`Gemini` includes chatbots and other AI tools
A partnership with Google would catapult Apple into the growing AI arms race
Deception
(50%)
The article is deceptive in several ways. Firstly, the author claims that Apple has developed a family of multimodal models called MM1 which can interpret and generate different types of data such as text and images at the same time. However, there is no evidence provided to support this claim.
The article states that 'Apple researchers say they've developed a family of multimodal models called MM1.'
However, there is no mention or link to any research paper or study which supports the development of these models.
Fallacies
(100%)
None Found At Time Of
Publication
Bias
(85%)
The article contains examples of both religious and monetary bias. The author uses language that dehumanizes those who hold certain beliefs about AI, such as calling them 'white supremacists' online celebrating the reference to a racist conspiracy theory. Additionally, the author mentions Google's Gemini AI engine in relation to Apple potentially partnering with them for their own AI efforts. This could be seen as an example of monetary bias since it implies that money is more important than other factors such as ethics or morality when making decisions about technology.
The author uses language that dehumanizes those who hold certain beliefs about AI, such as calling them 'white supremacists' online celebrating the reference to a racist conspiracy theory.
