🚀Apple's AI Breakthrough

Reading time: 5 minutes  

 Newsletter | News | Tools | Grow My Tool  

Today we will discuss-

• 👾Apple reveals new AI model 'MM1'

• 🎤Nvidia CEO addresses AGI, AI hallucinations

• 🙅‍♂️French employers hesitant to adopt AI

• ⚙️9 amazing AI tools you might not have heard of

All this and more - Let's dive in!

👩‍🍳What’s cooking in the newsroom?

Apple publishes details about its new multimodal AI model—MM1

🍎News - According to a recently published research paper, Apple researchers have developed a new generative AI model called MM1, which has the ability to process both text and visual data.

🤖What exactly can the model do? MM1, is a multimodal large language model (MLLM), which means it is trained on both images and text, enabling it to not only respond to text prompts but also tackle complex questions related to specific images.

The paper showcases MM1's ability to answer questions about photos and displaying general knowledge skills similar to chatbots like ChatGPT.

In an example from the Apple research paper, MM1 was presented with a photo of a sunlit restaurant table featuring beer bottles, along with an image of the menu. When prompted about the total cost of "all the beer on the table," the model accurately identified the price from the menu and calculated the total expense.

🔮What does MM1 mean for Apple's future? While it's difficult to make definitive conclusions about Apple's plans based solely on the research paper, the adaptability of multimodal models is evident. Therefore, it's reasonable to consider that MM1 may be a precursor to the development of a multimodal assistant capable of describing and answering questions about various types of media, such as photos, documents, or charts.

Nvidia's Jensen Huang says AI hallucinations are solvable, AGI is 5 years away

☕News - During a press conference at this week's annual GTC developer conference by Nvidia, CEO Jensen Huang shared his views on the trajectory of AGI, providing some interesting insights.

🤓Not another vague prediction?! Although certain press outlets persistently sought a timeline for the development of AGI, Huang stressed the difficulty in defining AGI and warned against sensationalist speculation. Huang argued that predicting when we'll see a functional AGI depends on how we define AGI. 

Drawing parallels to tangible milestones like New Year's Day or reaching a destination, Huang emphasized the significance of reaching a consensus on the criteria used to measure AGI achievement. 

Offering a nuanced perspective, Huang proposed practical benchmarks for AGI achievement. He suggested that if AGI were defined by specific criteria, such as outperforming humans by a considerable margin in tasks like legal bar exams, logic tests, or pre-med exams, it could be achievable within five years. However, he emphasized the necessity of clarity in defining AGI's parameters for accurate predictions.

😀What's more? When questioned about AI hallucinations—instances where AIs provide answers that sound plausible but lack factual basis—Huang remarked that the issue was easily solvable and recommended a research-based solution.

He proposed implementing a "retrieval-augmented generation" strategy, similar to basic media literacy, where AI verifies answers against reliable sources before responding. In critical areas like health advice, Huang advised cross-referencing multiple sources to ensure accuracy. In practical use, this would imply that AIs should also have the capability to acknowledge when they lack the necessary information.

More than 50% of French employers not using AI, don't plan to in the future

🙅‍♂️News - A recent poll by France's job search agency Pôle Emploi found that more than half of French employers are not using artificial intelligence (AI) and do not plan to do so in the future.

😧But why is that the case?

• Among those who do not use AI, 78% said their activity was incompatible with the technology.

• A further one-third of employers who do not use AI said they lack the skills to use it, while nearly one in three said they do not have the financial means to invest in it.

• Approximately 15% of employers who are not using AI said they were afraid of the technology.

🤔What other information did the report provide?

• Only 35% of employers reported that they were either already using AI or in the process of rolling it out in the workplace.

• The poll revealed that companies with over 200 employees were more prone to using the technology.

• Employers who used AI were very positive about its impact. In fact, 74% confessed that because of AI usage, there was a positive effect on the evolution of people's skill sets.

🙆🏻‍♀️What else is happening?

• Google hit with $270M fine in France as authority finds news publishers’ data was used for Gemini

• White House proposes up to $8.5B to fund Intel’s domestic chip manufacturing

• Microsoft hires DeepMind co-founder Mustafa Suleyman as CEO of new AI unit

• Saudi Arabia plans $40 billion push into Artificial Intelligence

• U.S. DOJ to target pre-IPO artificial intelligence frauds, top attorney says

• Segway brings a sub-$1,000 smart robot mower to the US

🗓️Historic moments

• 1927 - The German science fiction film Metropolis, directed by Fritz Lang, featured one of the first on-screen depictions of a robot. The film's robot character, Maria, has since become an iconic figure in science fiction, influencing numerous films, books, and other media.

• 1929 - Japanese biologist and professor Makoto Nishimura created Gakutensoku, the first robot to be built in Japan. This achievement marked not only Japan's first foray into robotics but also one of the earliest examples of humanoid robots globally, highlighting Japan's early contributions to the field.

• 1952 - American computer scientist Arthur Samuel developed a checkers-playing program that could learn to play better by itself. The Samuel Checkers-playing Program was among the world's first successful self-learning programs, and as such a very early demonstration of the fundamental concept of AI. 

👩🏼‍🚒Discover mind-blowing AI tools

1. Slazzer - An AI-powered tool that removes backgrounds from images quickly and automatically ($13.24/month) 

2. Surfer - Helps with SEO content creation and optimization; offers features like generating article outlines, providing keyword suggestions ($89/month) 

3. Questgen - Uses AI to automatically generate multiple-choice questions from any given text ($15/month)

4. TuneBlades - An AI-powered music editing tool that allows users to resize music tracks to any desired duration ($0.99/track)

5. Siwalu - AI-based animal recognition apps, including Dog Scanner, Cat Scanner, and Horse Scanner (Free)

6. Chord ai - Uses AI technology to provide chords, beats, and other musical information for any song (Free, $9/month)

7. Hexometer - A website monitoring tool that works 24/7 to detect and resolve issues before they impact businesses ($12/month)

8. Deepgram ASR - A speech-to-text tool that converts spoken language into written text with high accuracy, and speed (pay-as-you-go starting from $0.0020/min)

9. Supademo - An interactive demo platform that helps companies create engaging demos and guides (Free, $27/month) 

👨‍💻AI life hacks

• Use AI-generated music to enhance your creative projects - AI-powered tools like MuseNet and AIVA can generate original music and sound effects, perfect for adding depth and emotion to your videos, presentations, and other creative projects.

• Get motivated using AI - Feeling nervous before a big event is normal. But instead of seeking advice from others, you can turn to  AI chatbots for a personalized pre-event motivation strategy. 

• AI meeting assistants - If you've ever had to review meeting minutes and compose a summary email, you know it can be time-consuming and tedious. AI meeting assistants can help by seamlessly capturing your meetings' audio, video, and slides, efficiently generating summaries, allocating action items, and crafting follow-up emails.