Exploring Meta’s AI development goals and the involvement of Mark Zuckerberg and top executives in driving progress.
Meta Reveals AI Tool to Help Brands Target Facebook, Instagram Users
Meta, the parent company of Facebook and Instagram, plans to use its proprietary generative AI to create ads for brands on its social platforms by the end of the year. The AI tool will allow brands to create an array of advertising variations to target different types of users. Meta expects generative AI to eventually be used to create virtual worlds in the metaverse. The company has a large research organization with hundreds of employees focused on AI development. This move by Meta could lead to better and more cost-effective results for companies advertising on its platforms.
Meta Shares AI Model That Can Detect Objects It Hasn’t Seen Before.
Social media giant, Meta, has developed an AI model called “Segment Anything” that can detect objects in pictures and videos, even if they weren’t part of the training set. The technology can help reconstruct an object in 3D and limit the need for additional AI training. The model and dataset will be downloadable with a non-commercial license. However, the existing model is flawed and may miss finer details. The AI tool may help social networks keep up with a rapidly growing volume of content, and Meta wants to generalize computer vision. The company is under pressure to compete with tech heavyweights like Google and Microsoft, and it’s already planning generative AI “personas” for its social apps.
Of Course Facebook Will Monetize An Ad-Generating AI
Meta, formerly known as Facebook, is developing its own generative AI system capable of producing online ads and plans to charge businesses for its use. The technology is expected to become available in 2023, and the company’s chief technology officer, Andrew Bosworth, said that Meta’s generative AI software can help advertisers create marketing materials for specific audiences, saving them time and money. The social media giant, which makes most of its revenue from advertising, has not previously commercialized its own AI technology directly for other companies to use.
READ THE ARTICLE ON THE REGISTER.
Mark Zuckerberg is Spending Most of His Time on AI, says Meta CTO
Mark Zuckerberg and other top executives at Meta are spending most of their time working on artificial intelligence (AI), according to Meta’s Chief Technology Officer Andrew Bosworth. The company, which owns Facebook and Instagram, announced a product group in February focused on generative AI, and has its own large language model called LLaMa. Bosworth expects the company will debut some commercial applications using AI this year, including in its profit-driving ad business. Meta also hopes to eventually apply its AI technology to all of its services and products.
Has Amazon Lost The AI Race?
Amazon Web Services (AWS) has launched a 10-week AWS Generative AI Accelerator to boost its AI efforts. The program will offer access to AI resources, mentorship, and $300,000 in AWS credits. Though Amazon’s AI-powered digital assistant Alexa may have fallen short of creating the next big platform in tech, experts say Amazon’s real strength lies in the suite of AWS applications that power a large part of the internet. By investing in user-friendly platforms for developers to build and train more refined and specific AI models, Amazon will become a platform and unlock billions of dollars in value over the next decade.
Midjourney’s AI Platform Blocks Images of Chinese President, Sparking Ethics Debate
Midjourney, a San Francisco-based generative AI platform, has sparked an ethics debate by blocking the creation of images of Chinese President Xi Jinping. Critics argue that the move constitutes censorship and undermines free speech and expression. Midjourney CEO David Holz has previously stated that the company wants to “minimize drama” and suggested that the greater good is allowing Chinese people to use the technology rather than encouraging political satire. However, Senior Scholar Sarah McLaughlin from the Foundation for Individual Rights and Expression (FIRE) warns that such restrictions could lead to unintended consequences and follow other countries’ laws instead of US laws.
AI Fears Grow as Germany, France, Sweden, and Canada Voice Concerns
Governments around the world, including Canada, Germany, France, and Sweden, are expressing concerns about the potential risks posed by advanced artificial intelligence. Italy has even banned OpenAI’s ChatGPT chatbot after a bug exposed users’ payment information and chat history. However, some experts suggest that banning AI may not be realistic and that the focus should be on societal rules and economic policies that apply universally. OpenAI has attempted to address concerns about AI safety and privacy in a blog post but some critics argue that it is simply PR window dressing.
Doctors are Drowning in Paperwork. Some Companies Claim AI Can Help
Several startup companies are using AI chatbots to help reduce the amount of paperwork physicians have to complete, allowing them to spend more time with patients. Glass Health, a small San Francisco-based startup, has created “Glass AI,” a program that doctors can use to input patient information and receive a list of possible diagnoses and treatment plans. While some researchers are skeptical about the integration of AI into medicine, the technology is already present in healthcare, such as the use of an AI system to screen patients for diabetic retinopathy.
Here’s how AI and Virtual Reality can help Companies Boost Employee Engagement and Productivity
Virtual Reality (VR) and Artificial Intelligence (AI) are being used by companies to train employees and address issues such as disengagement and lowered productivity, according to Tacy Byham, CEO of DDI, an international human resources and leadership development consultancy. VR can be used for soft-skills training and provides a safe space for employees to practice conversations, allowing them to springboard into relevant and deep conversations. Incorporating VR and AI into a business strategy can benefit all levels of a company, but it is not a one-step solution, she added.
Want a Job in AI? These are the Skills You Need
The demand for AI jobs in the US is at its highest since 2014, and employers are looking for workers with different specific skill sets. According to Stanford University’s AI Index Report, the most in-demand skill today is machine learning, which teaches a system to make a prediction based on the data it’s trained on. Other AI skills that are in demand include AI, natural language processing, neural networks, autonomous driving, visual image recognition, and robotics. If you’re looking to pursue a career in AI, focusing on machine learning may be the most beneficial.
There Is Only One Question That Matters With AI
The Future of Life Institute has circulated a petition signed by nearly 3,000 people calling for a six-month moratorium on large-scale experiments with AI. Those who have signed the petition believe that the developers of GPT-4 and other large language model AIs have not taken appropriate steps to protect civilization from harm. They note that AI is just code where the benefits fall far short of what the promoters promise, with much more harm. The industry has long treated self-regulation as a license to do what it pleases. The question we should be asking about artificial intelligence and every other new technology is whether private corporations should be allowed to run uncontrolled experiments on the entire population without any guardrails or safety nets.
India Opts Against AI Regulation
India has decided against regulating AI’s growth in the country despite increasing calls for government intervention in the rapidly advancing technology. The Ministry of Electronics and IT stated that it has assessed the ethical concerns and risks of bias and discrimination associated with AI and is implementing necessary policies and infrastructure measures to cultivate a robust AI sector in the country. India’s decision, impacting the world’s second-largest internet market, comes as tech entrepreneurs and academics increasingly call for government intervention in AI, while few US lawmakers are actively pursuing regulation.
READ THE ARTICLE ON TECHCRUNCH.
AI Videos Are Freaky and Weird Now. But Where Are They Headed?
AI-generated videos are currently jarring, but their potential is massive. Tech companies are investing in text-to-video AI generators that can help cast movies, model scenes, and swap actors in and out of scenes. It may take years for generators to produce a full-length film based on prompts, but AI has the potential to revolutionize the entertainment industry. However, there are technical hurdles to overcome, and it is unlikely that AI will entirely replace human actors, directors, and writers anytime soon. Furthermore, the success of AI in film will depend on its ability to recreate human emotion and movie magic, rather than just generating distorted and demeaning content.
The Social Consequences of Using AI in Conversations
Using AI-enabled chat tools in conversations leads to more effective conversations and positive language use, according to researchers from Cornell University. The study found that participants perceived their chat partners more positively and used more positive language when using an AI platform called “Moshi,” which generated smart replies based on large language models. However, the group also found that when participants thought their partner was using AI-generated responses, they perceived that partner as less cooperative and felt less affiliation towards them. The study suggests that while AI can enhance communication efficiency and interpersonal perceptions, its overt use may undermine its potential benefits due to the negative connotations around AI.
READ THE ARTICLE ON NEUROSCIENCE NEWS.
Interpreting The Impact of AI Large Language Models on Chemistry
The emergence of improved ‘large language models’ (LLMs) has led to some researchers claiming that LLMs go beyond traditional deep-learning AI methods and display emergent features of the human mind, whilst others argue that they are merely exercises in finding correlations and are devoid of any kind of semantic understanding. However, experts suggest a third possibility, that LLMs possess a genuine kind of understanding that is quite distinct from that of the human mind. Recently, a paper was released that applied a LLM to the protein-structure algorithm, which produced promising results for the scientific industry.