AI Archives - Page 9 of 19 - Jet Developers Blog

Artificial Intelligence: Understanding the Future of AI

Artificial Intelligence (AI) has come a long way in the past few decades, and we now live in a world filled with exciting AI technologies.

Specialized algorithms and machine learning techniques have been developed to process vast amounts of data and make predictions based on patterns. We have also seen the emergence of AI chatbots like ChatGPT, smart home devices, virtual assistants like Siri, Google Assistants, and many more.

But here’s the thing: AI is still pretty limited. It can only do what we humans tell it to do, and it’s not great at handling tasks it hasn’t seen before.

That’s where artificial general intelligence (AGI) would come in – it would be like the superstar of the AI world. AGI would be the type of AI that can learn and reason like we humans do, which means it would have the potential to solve complex problems and make decisions independently.

Imagine having an AI system that can actually figure things out independently – now that’s something worth getting excited about!

While AGI is still in its early stages of development, it has the potential to revolutionize numerous industries, including healthcare, finance, transportation, and manufacturing. With AGI, medical research could lead to more accurate diagnoses and personalized treatments, while transportation systems could become more efficient and safer, leading to fewer accidents and less road congestion.

In this article, we will delve into the fascinating world of artificial general intelligence. We’ll explore its history, its potential impact on society, and the ethical and regulatory implications of its use.

What is artificial general intelligence (AGI)?

Artificial general intelligence (AGI) is a theoretical form of AI that can learn and reason like humans, potentially solving complex problems and making decisions independently. However, definitions of AGI vary as there is no agreed-upon definition of human intelligence. Experts from different fields define human intelligence from different perspectives.

However, those working on the development of AGI aim to replicate the cognitive abilities of human beings, including perception, understanding, learning, and reasoning, across a broad range of domains.

Unlike other forms of AI, such as narrow or weak AI, which are designed to perform specific tasks, AGI would perform a wide range of tasks, adapt to new situations, and learn from experience. AGI would reason about the world, form abstract concepts, and generalize knowledge from one domain to another. In essence, AGI would behave like humans without being explicitly programmed to do so.

Here are some of the key characteristics that would make AGI so powerful:

Access to vast amounts of background knowledge: AGI would tap into an extensive pool of knowledge on virtually any topic. This information would allow it to learn, adapt quickly, and make informed decisions.
Common sense: AGI would understand the nuances of everyday situations and respond accordingly. It could reason through scenarios that have not been explicitly programmed and use common sense to guide its actions.
Transfer learning: AGI could transfer knowledge and skills learned from one task to other related tasks.
Abstract thinking: AGI could comprehend and work with abstract ideas, enabling it to tackle complex problems and develop innovative solutions.
Understanding of cause and effect: AGI would be able to anticipate the outcomes of its decisions and take proactive measures to achieve its goals by understanding and using cause-and-effect relationships. This means that it could predict the consequences of its decisions and take proactive measures to achieve its goals.

The main difference between AGI and other forms of AI is the scope of their capabilities. While other forms of AI are designed to perform specific tasks, AGI would have the potential to perform a wide range of tasks, similar to humans.

The history of AGI

The quest for AGI has been a long and winding road. It began in the mid-1950s when the early pioneers of AI were brimming with optimism about the prospect of machines being able to think like humans. They believed that AGI was possible and would exist within a few decades. However, they soon discovered that the project was much more complicated than they had anticipated.

During the early years of AGI research, there was a palpable sense of excitement. Herbert A. Simon, one of the leading AI researchers of the time, famously predicted in 1965 that machines would be capable of doing any work a human can do within twenty years. This bold claim inspired the creation of the infamous character HAL 9000 in Arthur C. Clarke’s sci-fi classic 2001: A Space Odyssey (and the movie version by Stanley Kubrick).

However, the optimism of the early years was short-lived. By the early 1970s, it had become evident that researchers had underestimated the complexity of the AGI project.

Funding agencies became increasingly skeptical of AGI, and researchers were pressured to produce useful “applied AI” systems. As a result, AI researchers shifted their focus to specific sub-problems where AI could produce verifiable results and commercial applications.

Although AGI research was put on the back burner for several decades, it resurfaced in the late 1990s when Mark Gubrud used the term “artificial general intelligence” to discuss the implications of fully automated military production and operations. Around 2002, Shane Legg and Ben Goertzel reintroduced and popularized the term.

Despite renewed interest in AGI, many AI researchers today claim that intelligence is too complex to be completely replicated in the short term. Consequently, most AI research focuses on narrow AI systems widely used in the technology industry. However, a few computer scientists remain actively engaged in AGI research, and they contribute to a series of AGI conferences.

The potential impact of AGI

Picture this: a world where machines can solve some of the most complex problems, from climate change to cancer. A world where we no longer have to worry about repetitive, menial tasks because intelligent machines take care of them and many higher-level tasks. This, and more, is the potential impact of AGI.

The benefits and opportunities of AGI are endless. With its ability to process large amounts of data and find patterns, AGI could help us solve problems that have long baffled us. For instance, it could help us develop new drugs and treatments for chronic diseases like cancer. It could also help us better understand the complexities of climate change and find new ways to mitigate its effects.

AGI could also improve human life in countless ways. Automating tedious and dangerous tasks could free up our time and resources to focus on more creative and fulfilling pursuits. It could also revolutionize industries such as transportation and logistics by making them more efficient and safer. In short, AGI can change our lives and work in ways we can’t imagine.

However, there are also risks and challenges associated with the development of AGI. One of the biggest concerns is the displacement of jobs, as machines take over tasks previously done by humans. This could lead to economic disruption and social unrest – or a world where the only jobs left were either very high-level or menial jobs requiring physical labor. There are also significant ethical concerns, such as the possibility of machine bias in decision-making and the potential for misuse of AGI by those with malicious intent.

Public figures, including Elon Musk, Steve Wozniak, and Stephen Hawking, have endorsed the view that AI poses an existential risk for humanity. Similarly, AI researchers like Stuart J. Russell, Roman Yampolskiy, and Alexey Turchin support the basic thesis of AI’s potential threat to humanity.

Sharon Zhou, the co-founder of a generative AI company, believes that AGI is advancing faster than we can process, and we must consider how we use this powerful technology.

There are also safety risks associated with AGI, particularly if it becomes more advanced than human intelligence. Such machines could potentially be dangerous if they develop goals incompatible with human values. For example, if it’s given the task of combating global warming and it decides the best way is to eliminate the cause – humans.

Therefore, it’s essential to approach AGI development cautiously and establish proper regulations and safeguards to mitigate these risks.

The ethics of AGI

As artificial general intelligence (AGI) continues to make strides, it’s becoming increasingly important to consider the ethical implications of this technology. One of the primary concerns is whether or not AGI can learn and understand human ethics.

One worry is that if AGI is left unchecked, machines may make decisions that conflict with human values, morals, and interests. To avoid such issues, researchers must train the system to prioritize human life, understand and explain moral behavior, and respect individual rights and privacy.

Another ethical concern with AGI is the potential for bias in decision-making. If the data sets used to train AGI systems are biased, the resulting decisions and actions may also be biased, leading to unfair treatment or discrimination. We are already seeing this with weak AI. Therefore, ensuring that the data sets used to train AGI are diverse, representative, and free from bias is crucial.

Furthermore, there is the issue of responsibility and accountability. Who will be held accountable if AGI makes a decision that harms humans or the environment? Establishing clear guidelines and regulations for developing and using AGI is crucial to ensure accountability and responsibility.

The issue of job displacement is another concern with AGI. As AI becomes more intelligent, it will take over tasks previously done by humans, leading to job displacement and economic disruption.

Regulation and governance will play a critical role in ensuring responsible AI. Governments and organizations must work together now to establish ethical guidelines and standards for the development and use of AGI. This includes creating mechanisms for accountability and transparency in machine decision-making, ensuring that AGI is developed unbiased and ethically, and establishing safeguards to protect human safety, jobs, and well-being.

The future of AGI

The future of AGI development is a topic of much debate and speculation among experts in the field. While some believe that AGI is inevitable and will arrive sooner rather than later, others are skeptical about the possibility of ever achieving true AGI.

One potential outcome of AGI development is the creation of Artificial Super Intelligence (ASI), which refers to an AI system capable of surpassing human intelligence in all areas. Some experts believe that once AGI systems learn self-improvement, they can operate at a rate humans cannot control, leading to the eventual development of ASI.

However, there are concerns about the potential implications of ASI for society and the workforce. English physicist and author Stephen Hawking warned of the dangers of developing full artificial intelligence, stating that it could spell the end of the human race, as machines would eventually redesign themselves at an ever-increasing rate, leaving humans unable to compete.

Some experts, like inventor and futurist Ray Kurzweil, believe that computers will achieve human levels of intelligence soon (Kurzweil believes this will be by 2029) and that AI will then continue to improve exponentially, leading to breakthroughs that enable it to operate at levels beyond human comprehension and control.

Recent developments in generative AI have brought us closer to realizing the vision of AGI. User-friendly generative AI interfaces like ChatGPT have demonstrated impressive capabilities to understand human text prompts and answer questions on a limitless range of topics, although this is still all based on interpreting data that has been produced by humans. Image generation systems like DALL-E have also upended the visual landscape, generating realistic images just from a scene description, again, based on work by humans.

Despite these developments, AGI’s limitations and dangers are already well-known among users. As a result, AGI development will likely continue to be a hotly debated topic, with significant implications for the future of work and society.

Conclusion

Artificial general intelligence (AGI) can potentially revolutionize the world as we know it. From advancements in medicine to space exploration and beyond, AGI could solve some of humanity’s most pressing problems.

However, the development and deployment of AGI must be approached with caution and responsibility. We must ensure that these systems are aligned with human values and interests and do not threaten our safety and well-being.

With continued research and collaboration among experts in various fields, we can strive towards a future where AGI benefits society while mitigating potential risks.

The future of AGI is an exciting and rapidly evolving field, and it is up to us to shape it in a way that serves humanity’s best interests.

Meta Trains AI Models on the Bible to Learn Over 1,000 Languages

Meta, a leading tech company, has developed new AI models that were trained using the Bible to recognize and generate speech in over 1,000 languages. The company aims to employ these algorithms in efforts to preserve languages that are at risk of disappearing.

Currently, there are approximately 7,000 languages spoken worldwide. To empower developers working with various languages, Meta is making its language models publicly available through GitHub, a popular code hosting service. This move encourages the creation of diverse and innovative speech applications.

The newly developed models were trained on two distinct datasets. The first dataset contains audio recordings of the New Testament Bible in 1,107 languages, while the second dataset comprises unlabeled New Testament audio recordings in 3,809 languages. By leveraging these comprehensive datasets, Meta’s research scientist, Michael Auli, explains that the models can be utilized to build speech systems with minimal data.

While languages like English possess extensive and reliable datasets, the same cannot be said for smaller languages spoken by limited populations, such as those spoken by only 1,000 individuals. Meta’s language models provide a solution to this data scarcity, enabling the development of speech applications for languages lacking adequate resources.

The researchers assert that their models can not only converse in over 1,000 languages but also recognize more than 4,000 languages. Furthermore, when compared to rival models like OpenAI Whisper, Meta’s version exhibited a significantly lower error rate despite covering a broader range of languages, exceeding even 11 times more language coverage.

However, the scientists acknowledge that the models may occasionally mistranscribe specific words or phrases. Additionally, their speech recognition models displayed a slightly higher occurrence of biased words compared to other models, albeit only by a marginal increase of 0.7%.

Chris Emezue, a researcher at Masakhane, an organization focused on natural-language processing for African languages, expressed concerns about the use of religious text, such as the Bible, as the basis for training these models. He believes that the Bible carries inherent biases and misrepresentations, which could impact the accuracy and neutrality of the models’ outputs.

This development poses an important question: Is Meta’s advancement in language models a step forward, or does its utilization of religious text for training introduce controversial elements that hinder its overall impact? The conversation around the ethical considerations and potential biases involved in training language models remains ongoing.

Regulators Turn to Old Laws to Tackle AI Technology like ChatGPT

Organizations like the European Union (EU) are taking the lead in formulating new regulations for AI, which could potentially establish a global standard. However, the enforcement of these regulations is expected to be a time-consuming process that spans several years.

“In the absence of specific regulations, governments can only resort to the application of existing rules,” stated Massimiliano Cimnaghi, a European data governance expert at consultancy BIP, in a statement to Reuters.

As a result, regulators are turning to already-established laws, such as data protection regulations and safety measures, to tackle concerns related to personal data protection and public safety. The necessity for regulation became evident when national privacy watchdogs across Europe, including the Italian regulator Garante, took action against OpenAI’s ChatGPT, accusing the company of violating the EU’s General Data Protection Regulation (GDPR).

In response, OpenAI implemented age verification features and provided European users with the ability to block their data from being used to train the AI model.

However, this incident prompted additional data protection authorities in France and Spain to initiate investigations into OpenAI’s compliance with privacy laws.

Consequently, regulators are striving to apply existing rules that encompass various aspects, including copyright, data privacy, the data utilized to train AI models, and the content generated by these models.

Proposals for the AI Act

In the European Union, proposals for the AI Act will require companies like OpenAI to disclose any copyrighted material used to train their models, exposing them to potential legal challenges. However, proving copyright infringement may not be straightforward, as Sergey Lagodinsky, a politician involved in drafting the EU proposals, explains.

“It’s like reading hundreds of novels before you write your own,” he said. “If you actually copy something and publish it, that’s one thing. But if you’re not directly plagiarizing someone else’s material, it doesn’t matter what you trained yourself on.

Regulators are now urged to “interpret and reinterpret their mandates,” says Suresh Venkatasubramanian, a former technology advisor to the White House. For instance, the U.S. Federal Trade Commission (FTC) has used its existing regulatory powers to investigate algorithms for discriminatory practices.

Similarly, French data regulator CNIL has started exploring how existing laws might apply to AI, considering provisions of the GDPR that protect individuals from automated decision-making.

As regulators adapt to the rapid pace of technological advances, some industry insiders call for increased engagement between regulators and corporate leaders.

Harry Borovick, general counsel at Luminance, a startup that utilizes AI to process legal documents, expresses concern over the limited dialogue between regulators and companies.

He believes that regulators should implement approaches that strike the right balance between consumer protection and business growth, as the future hinges on this cooperation.

While the development of regulations to govern generative AI is a complex task, regulators worldwide are taking steps to ensure the responsible use of this transformative technology.

Sci-Fi Author Says He Wrote 97 Books Using AI Tools

Sci-fi author Tim Boucher recently shared his remarkable achievement of creating 97 books within a span of nine months, thanks to the assistance of artificial intelligence (AI). In an article published by Newsweek, Boucher revealed that he utilized various AI tools to bring his vision to life.

Boucher employed the AI image generator called Midjourney to illustrate the contents of his books. For brainstorming and text generation, he relied on ChatGPT and Anthropic’s Claude. Each of his novels ranged from 2,000 to 5,000 words and included an impressive 40 to 140 AI-generated images. The author mentioned that, on average, it took him approximately six to eight hours to create and publish a book using AI tools, although some could be completed in as little as three hours.

To make his creations available to the public, Boucher opted to sell his books online, with prices ranging from $1.99 to $3.99 per copy. In his article for Newsweek, he expressed his appreciation for the role AI played in his creative process, noting that it significantly boosted his productivity while maintaining consistent quality. Furthermore, AI tools enabled him to delve into intricate world-building with unparalleled efficiency.

The market has recently witnessed a surge in AI-generated novels. In February, ChatGPT was attributed as the author or coauthor of over 200 titles listed in Amazon’s bookstore. The genres that garnered the most attention were AI guides and children’s books.

One example of AI-assisted book creation comes from Ammaar Reshi, a product-design manager at a San Francisco-based financial-tech company. Reshi revealed that he utilized ChatGPT and Midjourney to write and illustrate a children’s book titled “Alice and Sparkle” in just 72 hours. However, the book stirred controversy on Twitter, as it faced backlash from creatives. Concerns were raised regarding the use of AI image generators and the perceived quality of the writing.

The advent of AI in the realm of book creation has opened up new possibilities, allowing authors like Tim Boucher to expand their creative output. While these developments have sparked debates within the artistic community, it is undeniable that AI continues to reshape the landscape of literary expression.

Artists Deserve ‘Fairness and Control’ Over AI Use, Says SoundExchange CEO

During OpenAI CEO Sam Altman’s testimony before a Senate panel, many became acquainted with SoundExchange, a Washington, DC-based nonprofit organization established two decades ago to collect royalties from digital music platforms and distribute them to music creators.

Senator Marsha Blackburn (R-TN) questioned Altman regarding the compensation of songwriters and musicians when their works are utilized by AI companies. She emphasized that the Nashville music community should have the authority to decide if their copyrighted songs and images are used for training these AI models. Blackburn inquired whether a system similar to SoundExchange could be employed for the collection and distribution of funds to compensate artists.

Although Altman claimed to be unfamiliar with SoundExchange, he acknowledged the importance of ensuring that content creators benefit from AI technology.

Michael Huppe, President and CEO of SoundExchange, as well as an adjunct professor in music law at Georgetown University, expressed his satisfaction with Blackburn’s remarks. He acknowledged the rapidly evolving landscape, where AI-generated songs mimicking popular artists can go viral, platforms like Grimes’ enable anyone to create AI-generated songs using their voice, and AI is used to release songs featuring deceased artists like Notorious B.I.G.

Huppe commended Senator Blackburn for her forward-thinking approach in recognizing the necessity of allowing the creative class to actively participate and be fairly compensated within this new technological landscape. He emphasized that AI is here to stay, underscoring the importance of compensating and protecting the work of artists in this realm.

Not just about artists — even the NFL is concerned

How AI development affects creative workers is not just about the music industry, Huppe emphasized. He pointed to the March launch of the Human Artistry Campaign, a set of principles that outline the responsible use of AI to “support human creativity and accomplishment with respect to the inimitable value of human artistry and expression.” The campaign, he said, has been joined by over 100 organizations representing songwriters, musicians, authors, literary agents, publishers, voice actors and photographers — as well as non-artistic entities like sports organizations, including the Major League Baseball Players Association and the NFL Players Association.

Why sports? “Many players profit off their name, image and likeness,” said Huppe. “So this isn’t just about copyright when we talk about what happens [with AI]. It’s also how generative AI — whether text, images, audio or video — can capitalize on those who have built up their brand and persona. You have someone else trying to capitalize on that without permission.”

Creative class “getting louder” about AI

The bottom line, Huppe said, is that how AI uses creators’ work should be their choice. “It’s about fairness and control, so that the creative class can’t just have these things taken away from them.”

Huppe pointed out that there is already a nascent marketplace developing of people licensing their works for AI, such as how OpenAI licensed images from Shutterstock to train its models. “You can imagine a world where that starts to be the norm,” he said, “where there’s an organized licensing structure and ethical AI companies can know what’s allowed to be scraped and what’s off-limits … and where they share part of their profits with the creative community.”

With other industries pushing back on generative AI — including lawsuits filed by visual artists, striking Hollywood writers and unionizing journalists — and celebrities like Justine Bateman and Sting speaking out, Huppe said the creative class “is getting louder as we speak.”

Music, he said, has often been like “the marines on the beach” when it comes to dealing with new technologies that ultimately affect all industries: “There’s almost no industry that doesn’t have the risk of being really impacted by generative AI. It’s on everybody’s mind.”

AI-Powered Automation Enhances Job Fulfillment for Nearly 60% of Workers: Report

According to a recent survey conducted by automation software firm UiPath, a substantial majority of workers (approximately 60%) believe that AI-powered automation solutions can mitigate burnout and significantly improve job satisfaction. Moreover, 57% of the respondents expressed a more positive perception of employers that integrate business automation to support their employees and streamline operations than of employers that do not, reflecting their favorable attitude towards such practices.

As workloads intensify, 28% of individuals report taking on extra responsibilities due to layoffs or hiring freezes. A full 29% of workers worldwide experience burnout. This is fueling an escalating dependence on AI tools for alleviation.

The automation generation

These factors are contributing to the emergence of what has been called the “automation generation” — professionals who proactively adopt automation and AI to enhance collaboration, foster creativity and boost productivity, regardless of age or demographic.

These individuals actively seek technologies that enhance their professional and personal lives, as they strive to avoid feeling dehumanized.

One of the survey’s primary revelations is that 31% of respondents actively employ business automation solutions in their workplaces.

The automation generation subgroup believes they have the resources and support they need (87%) to carry out their responsibilities effectively. Furthermore, 83% of these workers believe that business automation solutions can effectively mitigate burnout and enhance job satisfaction.

“With more than half of respondents stating they believe automation can address burnout and improve job fulfillment, it is clear that AI-powered business automation technology is already positively impacting business and technical workers and helping them to reduce time spent on repetitive tasks and focus on more critical and gratifying work,” Brigette McInnis-Day, chief people officer at UiPath.

She emphasized that this assertion is reinforced by the fact that among the respondents who are already using business automation solutions, 80% believe that these solutions enable them to perform their jobs more effectively, and 79% hold a more positive perception of employers that implement business automation than of those that don’t.

The survey, administered in March 2023, was conducted in partnership with Researchscape. It garnered online responses from 6,460 executives worldwide. Topline results were weighted to ensure representation of each country’s GDP, with the following distribution: U.S. (55%), Japan (10%), Germany (9%), India (8%), United Kingdom (7%), France (6%), Australia (4%) and Singapore (2%).

Tackling office workloads through AI

The survey reveals that workers worldwide are increasingly embracing automation and AI-powered tools to tackle mundane tasks.

Specifically, respondents expressed their desire for automation to assist in tasks such as data analysis (52%), data input/creation (50%), IT/technical issue resolution (49%) and report generation (48%).

When questioned about the sources of their burnout and work fatigue, respondents highlighted working beyond scheduled hours (40%), pressure from managers and leadership (32%), and excessive time dedicated to tactical tasks (27%) as the primary causes.

“AI-powered automation emerges as a solution to alleviate these leading causes of burnout, enabling workers to swiftly and effortlessly locate and analyze data while streamlining repetitive and time-consuming tasks,” McInnis-Day.

Workers of the automation generation emphasize flexibility, career advancement and focused work time. In terms of where automation tools impact their jobs, respondents expressed the desire for enhanced flexibility in their work environments (34%), allocated time for acquiring new skills (32%) and dedicated hours for critical tasks (27%).

“Unlike the previous defining generational categories, the automation generation encompasses all ages and demographics,” explained McInnis-Day. “It is the professionals embracing AI to be more collaborative, creative and productive as well as using these technologies to deliver more satisfying, positive workplace experiences, enrich their personal lives and prevent them from overall feeling like robots themselves. They are looking for a renewed and revived sense of purpose in their work — and automation is helping them realize that.”

Not surprisingly, the survey revealed that younger employees are more receptive to these new technologies. Majorities of Generation Z (69%), Millennials (63%) and Generation X (51%) respondents firmly believe that automation has the potential to enhance their job performance.

“Among the workers surveyed, 31% of respondents said they were already utilizing business automation solutions (of this group, 39% were Millennials and 42% were Gen Z). Additionally, of the 31% already using business automation solutions, 87% feel they have the resources and support needed to do their job effectively,” added McInnis-Day. “The findings prove that employees using AI-powered automation believe in its ability to advance their careers and support work-life balance.”

The growing demand for automation and AI-powered tools

According to McInnis-Day, persistent economic uncertainty and the need for organizations to accomplish more with fewer resources will drive increasing demand for automation and AI-powered tools. She said that companies that adopt an open and adaptable approach to deploying AI are best positioned to attract skilled employees who can contribute to their success.

“The top resource the automation generation identified as the key aspect that would help them do their jobs better and/or advance was technical tools and software,” she said. “Fifty-eight percent of respondents indicated they were looking for these technology tools to help them respond to today’s economic and labor market pressures.”

She advises business leaders to equip their workers with AI-powered automation tools to thrive in an automation-first world and alleviate resource constraints.

“These survey results provide compelling evidence that incorporating AI-powered automation across the organization is not only a wise investment but also aligns with employees’ preferences,” she said. “With workloads on the rise and employees seeking careers that offer a healthy work-life balance, the integration of AI-powered automation becomes crucial in delivering more fulfilling and positive workplace experiences.”

UN AI Adviser Warns About the Destructive Use of Deepfakes

Neil Sahota, an artificial intelligence (AI) expert and adviser to the United Nations, recently raised concerns about the increasing threat posed by highly realistic deepfakes. In an interview with CTVNews.ca on Friday, Sahota highlighted the risks associated with these manipulated media creations.

Sahota described deepfakes as digital replicas or mirror images of real-world individuals, often created without their consent and for malicious purposes, primarily aimed at deceiving or tricking others. The emergence of deepfakes has resulted in various instances of fake content going viral, encompassing a wide range of topics, including political simulations and celebrity endorsements.

While famous individuals have often been the primary targets, Sahota emphasized that ordinary civilians are also vulnerable to this form of manipulation. He noted that deepfakes initially gained traction through the distribution of revenge porn, highlighting the importance of remaining vigilant.

To identify manipulated media, Sahota advised individuals to pay attention to subtle inconsistencies in video and audio content. Signs to watch out for include unusual body language, odd shadowing effects, and discrepancies in the spoken words. By maintaining a vigilant eye and questioning the authenticity of media, individuals can become better equipped to identify potential deepfake content.

As deepfake technology continues to advance, Sahota’s warnings serve as a reminder of the critical need to exercise caution and skepticism when consuming digital media, as well as the urgent need for proactive measures to address the risks associated with deepfakes.

Not enough

Sahota also argued that currently policymakers are not doing enough in terms of educating the public on the many dangers of deepfakes and how to spot them. He recommended that a content verification system be implemented that would use digital tokens to authenticate media and identify deepfakes.

“Even celebrities are trying to figure out a way to create a trusted stamp, some sort of token or authentication system so that if you’re having any kind of non-in-person engagement, you have a way to verify,” he told CTVNews.ca.

“That’s kind of what’s starting to happen at the UN-level. Like, how do we authenticate conversations, authenticate video?”

Yellow AI Launches YellowG, A Generative AI Platform

Yellow AI, a leading conversational AI platform, has unveiled YellowG, a state-of-the-art conversational artificial intelligence (AI) platform specifically designed for automation technology. Leveraging generative AI and enterprise GPT capabilities, Yellow AI aims to empower enterprises across various industries by offering tailored solutions that streamline complex workflows, enhance existing processes, and foster innovation.

The platform features an advanced multi-large language model (LLM) architecture, continuously trained on billions of conversations. Yellow AI asserts that this architecture ensures exceptional scalability, speed, and accuracy, enabling businesses to unleash the full potential of the platform.

According to Yellow AI, integrating AI-driven chatbots like YellowG into customer and employee experiences across multiple channels can significantly elevate levels of automation for businesses. The company claims that such integration not only leads to substantial operational cost reductions but also enables up to 90% automation within the first 30 days.

“Our groundbreaking platform eliminates the need for setup time, providing instant usage as soon as a bot is built,” said Raghu Ravinutala, CEO and co-founder of Yellow AI, in an interview with VentureBeat. “With robust, enterprise-level security, it ensures maximum safety through a combination of centralized global and proprietary LLMs. Our real-time generative AI approach is specifically designed to drive enterprise conversations. This means YellowG can dynamically generate workflows and effortlessly handle complex scenarios.”

YellowG’s zero setup time, coupled with its robust security measures, positions it as a game-changer in the conversational AI landscape. By harnessing the power of generative AI, the platform empowers enterprises to automate processes, enhance customer interactions, and navigate intricate business scenarios with ease. Yellow AI is dedicated to delivering cutting-edge conversational AI solutions that drive efficiency, productivity, and growth for businesses of all sizes.

AI with human touch

The new tool empowers users to generate runtime workflows and make real-time decisions using dynamic AI agents, said Ravinutala. Moreover, it adds a unique human touch to AI conversations by demonstrating near-human empathy while maintaining an impressively low hallucination rate close to zero.

In addition to its multi-LLM architecture, YellowG utilizes enterprise data and industry-specific knowledge to navigate complex scenarios. The chatbot’s capacity to comprehend the context of conversations enables it to provide personalized responses that are finely tailored to specific use cases.

“The YellowG workflow generator is powered by the ‘dynamic AI agent,’ our orchestrator engine that harnesses the power of multiple LLMs,” said Ravinutala. “It utilizes knowledge from our proprietary platform data, the anonymized historical record of customer interactions and enterprise data.”

Yellow AI claims a response intent accuracy rate of more than 97%. In addition, the company asserts its capability to learn from extensive volumes of data, enabling it to generate responses to even the most intricate queries that traditional conversational AI platforms may find challenging.

Automating business workflows through generative AI

When a customer’s message enters the conversational interface, YellowG promptly analyzes it to decipher the request and develop a strategic plan for fulfilling their goal. Subsequently, generative AI interacts with the enterprise system to retrieve all relevant data necessary for processing the user’s request.

Leveraging this data, the platform utilizes an LLM orchestration layer to formulate and fine-tune the AI bot’s response. This ensures accurate alignment between the generated response, the obtained information and the customer’s initial request.

YellowG implements responsible AI practices during the post-processing stage by rigorously examining security, compliance and privacy measures. After that review, it delivers responses exhibiting human-like characteristics, showcasing exceptional accuracy and virtually no hallucinations.

“All the while, it remains focused on achieving the business objectives,” said Ravinutala. “Our multi-LLM architecture combines centralized LLMs’ intelligence with the precision and security of proprietary LLMs.”

Real-time generative AI

By integrating advanced AI and natural language processing (NLP) technologies, the platform provides customers with a human-like experience. The company said that the platform generates responses that are not pre-scripted by utilizing real-time generative AI, resulting in a more natural and seamless conversation flow.

“Our platform has been designed to detect and interpret the emotional tone and sentiment expressed in the customer’s message,” Ravinutala explained. “It can recognize various emotions such as frustration, confusion, happiness or the need for assistance, allowing it to adapt responses and provide the emotional support that one would typically expect from a human agent. This empathetic interaction establishes a deeper level of understanding, assuring customers that their sentiments are truly acknowledged.”

A prominent feature of YellowG is its capability to adapt to the customer’s unique communication style and requirements. For example, whether a customer prefers brief and concise answers or requires more comprehensive explanations, YellowG can adjust its responses accordingly.

The platform’s AI agent also leverages real-time analysis of the user’s responses to guide the conversation, resulting in highly personalized and tailored interaction.

Zero setup for instant LLM incorporation

YellowG’s zero setup feature empowers it to ingest and analyze its customers’ documents and websites. This comprehensive integration of knowledge enables the platform to deliver instant answers to any inquiries that fall within the scope of these resources.

“For customers with extensive knowledge repositories, this capability alone allows us to deliver a high level of automation from day one,” said Ravinutala.

Furthermore, the platform’s no-code solutions facilitate seamless connectivity with customer APIs, enabling the implementation of static workflows that unlock a new realm of automation. However, the company said it’s important to note that static workflows have limitations when handling fluid conversations, often imposing rigid conversational flows on users.

“To overcome this limitation, we have implemented dynamic runtime workflows that adapt based on user input,” Ravinutala added. “This approach empowers us to automate a significantly large number of customer queries.”

Ravinutala said the company has successfully developed proprietary data-trained LLMs in-house for various domains and use cases, including document Q&A, contextual history and summarization.

Yellow AI’s primary focus is tackling complex end-user-facing scenarios within customer support, marketing and employee experience where real-time decision-making is crucial. Ultimately, the goal is to leverage LLMs during runtime to redefine and enhance end-user experiences.

“One such use case that we solved using an in-house model is summarization for situations that demand fast response times,” he said. “We have also created a proprietary context model that empowers our dynamic AI agents to understand the conversation’s context more accurately.”

Safeguarding customer data through security compliance

According to the company, YellowG is engineered to be genuinely multi-cloud and multi-region, adhering to the most stringent security standards and compliance requirements. In addition, it implements rigorous measures to conceal Personally Identifiable Information (PII) from third-party LLMs, effectively safeguarding customer data.

Moreover, the platform successfully fulfills the criteria SOC 2 Type 2 certification sets forth. This certification attests to the fact that YellowG’s systems and processes are purposefully designed to protect customer data while maintaining exemplary levels of security and privacy.

“To enhance data access control, Yellow AI employs a role-based access control (RBAC) system, giving customers the ultimate authority to define access privileges,” said Ravinutala. “Every message exchanged through our platform is encrypted at rest using AES 256 encryption and in transit using TLS 1.2 and above.”

Empowering an AI-First Future: Meta Unveils New AI Data Centers and Supercomputer

Meta, formerly known as Facebook, has been at the forefront of artificial intelligence (AI) for over a decade, utilizing it to power their range of products and services, including News Feed, Facebook Ads, Messenger, and virtual reality. With the increasing demand for more advanced and scalable AI solutions, Meta recognizes the need for innovative and efficient AI infrastructure.

At the recent AI Infra @ Scale event, a virtual conference organized by Meta’s engineering and infrastructure teams, the company made several announcements regarding new hardware and software projects aimed at supporting the next generation of AI applications. The event featured Meta speakers who shared their valuable insights and experiences in building and deploying large-scale AI systems.

One significant announcement was the introduction of a new AI data center design optimized for both AI training and inference, the primary stages of developing and running AI models. These data centers will leverage Meta’s own silicon called the Meta training and inference accelerator (MTIA), a chip specifically designed to accelerate AI workloads across diverse domains, including computer vision, natural language processing, and recommendation systems.

Meta also unveiled the Research Supercluster (RSC), an AI supercomputer that integrates a staggering 16,000 GPUs. This supercomputer has been instrumental in training large language models (LLMs), such as the LLaMA project, which Meta had previously announced in February.

“We have been tirelessly building advanced AI infrastructure for years, and this ongoing work represents our commitment to enabling further advancements and more effective utilization of this technology across all aspects of our operations,” stated Meta CEO Mark Zuckerberg.

Meta’s dedication to advancing AI infrastructure demonstrates their long-term vision for utilizing cutting-edge technology and enhancing the application of AI in their products and services. As the demand for AI continues to evolve, Meta remains at the forefront, driving innovation and pushing the boundaries of what is possible in the field of artificial intelligence.

Building AI infrastructure is table stakes in 2023

Meta is far from being the only hyperscaler or large IT vendor that is thinking about purpose-built AI infrastructure. In November, Microsoft and Nvidia announced a partnership for an AI supercomputer in the cloud. The system benefits (not surprisingly) from Nvidia GPUs, connected with Nvidia’s Quantum 2 InfiniBand networking technology.

A few months later in February, IBM outlined details of its AI supercomputer, codenamed Vela. IBM’s system is using x86 silicon, alongside Nvidia GPUs and ethernet-based networking. Each node in the Vela system is packed with eight 80GB A100 GPUs. IBM’s goal is to build out new foundation models that can help serve enterprise AI needs.

Not to be outdone, Google has also jumped into the AI supercomputer race with an announcement on May 10. The Google system is using Nvidia GPUs along with custom designed infrastructure processing units (IPUs) to enable rapid data flow.

What Meta’s new AI inference accelerator brings to the table

Meta is now also jumping into the custom silicon space with its MTIA chip. Custom built AI inference chips are also not a new thing either. Google has been building out its tensor processing unit (TPU) for several years and Amazon has had its own AWS inferentia chips since 2018.

For Meta, the need for AI inference spans multiple aspects of its operations for its social media sites, including news feeds, ranking, content understanding and recommendations. In a video outlining the MTIA silicon, Meta research scientist for infrastructure Amin Firoozshahian commented that traditional CPUs are not designed to handle the inference demands from the applications that Meta runs. That’s why the company decided to build its own custom silicon.

“MTIA is a chip that is optimized for the workloads we care about and tailored specifically for those needs,” Firoozshahian said.

Meta is also a big user of the open source PyTorch machine learning (ML) framework, which it originally created. Since 2022, PyTorch has been under the governance of the Linux Foundation’s PyTorch Foundation effort. Part of the goal with MTIA is to have highly optimized silicon for running PyTorch workloads at Meta’s large scale.

The MTIA silicon is a 7nm (nanometer) process design and can provide up to 102.4 TOPS (Trillion Operations per Second). The MTIA is part of a highly integrated approach within Meta to optimize AI operations, including networking, data center optimization and power utilization.

Zoom Makes a Big Bet on AI with Investment in Anthropic

In a significant move towards harnessing the power of generative AI, Zoom has announced its collaboration with AI startup Anthropic. Building on their existing partnership with OpenAI, the enterprise communication company revealed its plans to integrate Anthropic’s innovative Claude AI assistant into Zoom’s productivity platform. To solidify this collaboration, Zoom’s global investment arm has also made an undisclosed financial investment in Anthropic, which is supported by Google.

This strategic partnership aligns with Zoom’s federated approach to AI and comes at a time when competitors like Microsoft are actively incorporating AI-driven capabilities into Teams, Google is integrating AI into Workspace, and Salesforce is focusing on SlackGPT.

However, Zoom has outlined its initial objective of incorporating Claude within its omni channel contact center offerings before expanding its integration to other segments of the platform. The specific timeline and execution details for the broader integration have not been disclosed at this time.

How exactly will Claude help in Zoom Contact Center?

Zoom’s Contact Center is a video-first support hub that improves customer support for enterprises. It includes multiple products, including Zoom Virtual Agent and Zoom Workforce Management.

With the Anthropic partnership, Zoom plans to integrate Claude across the entire Contact Center portfolio to build self-service features that not only improve end-user outcomes but also enable superior agent experiences.

For instance, it will be able to understand customers’ intent from their inputs and guide them to the best solution, and provide actionable insights that managers can use to coach agents.

“Anthropic’s Constitutional AI model is primed to provide safe and responsible integrations for our next-generation innovations, beginning with the Zoom Contact Center portfolio,” said Smita Hashim, chief product officer at Zoom. “With Claude guiding agents toward trustworthy resolutions and powering self-service for end users, companies will be able to take customer relationships to another level.”

Moving ahead, Zoom Contact Center will also use Claude to provide the right resources to agents, enabling them to deliver improved customer service, a company spokesperson told VentureBeat. They added that Claude’s capabilities will be expanded across the Zoom platform — which includes Team Chat, Meetings, Phone and Whiteboard — but did not share specific details.

The federated approach

The partnership with Anthropic is Zoom’s latest move in its federated approach to AI, where it is using its own proprietary AI models along with those from leading AI companies and select customers’ own models.

“With this flexibility to incorporate multiple types of models, our goal is to provide the most value for our customers’ diverse needs. These models are also customizable, so they can be tuned to a given company’s vocabulary and scenarios for better performance,” Hashim said in a blog post.

Zoom has already been working with OpenAI for IQ, its conversational intelligence product. In fact, back in March, Zoom announced multiple AI-powered capabilities for the product with OpenAI, including the ability to generate draft messages and emails and provide summaries for chat threads. The capabilities started rolling out for select customers in April.