Chat Gpt

Latest news headlines about artificial intelligence

Forget Chatbots. AI Agents Are the Future

March 14, 2024, 4 p.m. • WIRED • (4 Minute Read)

The report by WIRED reveals that technological advancements are shifting from chatbots to AI agents capable of executing tasks. Cognition AI's demonstration of the AI program, Devin, performing tasks customarily carried out by human software engineers, has drawn attention. Devin conceptualized solutions, generated and tested code, and even created a website to summarize results. The emergence of these AI agents, exemplified by technologies such as Auto-GPT and vimGPT, showcases the potential for these agents to take actions, but challenges persist with error rates and potential consequences. Notably, Google DeepMind has developed an agent, SIMA, that efficiently plays complex video games and may have broader applications in the future. The CEO of Google DeepMind, Demis Hassabis, anticipates significant advancements in these AI agents, marking a shift towards more capable and reliable systems.

Latest Prompt Engineering Trend Uses Generative AI To Generate Your Prompts For You

March 14, 2024, 11:15 a.m. • Forbes • (14 Minute Read)

The latest trend in prompt engineering is the use of generative AI to create prompts. Dr. Lance B. Eliot, a world-renowned expert on Artificial Intelligence (AI) and Machine Learning (ML), emphasizes the importance of leveraging generative AI to devise prompts for optimal results. He emphasizes the significance of well-crafted prompts in guiding the model to produce relevant and coherent outputs. Eliot provides a set of handy tips for using generative AI to generate personalized prompts, such as obtaining feedback on hand-crafted draft prompts and analyzing prompting strategies. He also shares examples of employing ChatGPT, a popular generative AI app, to generate and review prompts, highlighting the necessity of specific and clear prompts in guiding the AI to produce informative responses. Overall, the article serves as a comprehensive guide to leveraging generative AI for prompt engineering and underscores the critical role of well-devised prompts in optimizing AI interactions.

Regulators Need AI Expertise. They Can't Afford It

March 14, 2024, 11 a.m. • WIRED • (4 Minute Read)

The European AI Office and the UK government are struggling to attract AI expertise due to the low salaries being offered in comparison to industry compensation. As the AI boom continues, regulators are urgently seeking AI experts to study and regulate the fast-paced developments in AI. However, job ads offering salaries well below industry standards are causing concerns among AI researchers and engineers. The European AI Office and the UK government are offering salaries far lower than what is being offered within the industry, leading to a brain drain and potential challenges in attracting top talent. Despite efforts to recruit from the industry, the public sector is struggling to offer competitive compensation. This situation highlights the growing disparity between private and public sector compensation and raises questions about the ability of regulators to keep up with the pace of AI development.

Microsoft Copilot: Free tier gets AI upgrade to GPT-4 Turbo

March 14, 2024, 9 a.m. • Mashable • (3 Minute Read)

Microsoft Copilot's free tier is set to receive a major upgrade with the introduction of GPT-4 Turbo, a more advanced version of the AI model. The new upgrade brings several enhancements to the AI assistant, including an extended knowledge base cut-off date, an improved ability to process larger amounts of text in a single prompt, and a reduction in "laziness" that had been previously observed in the GPT-4 model. The upgrade to GPT-4 Turbo aims to provide free users with an enhanced and more capable Copilot experience. This development comes amid rumors of a new GPT-4.5 Turbo model, which is speculated to be exclusive to the paid tier.

OpenAI's deals with publishers could spell trouble for rivals

March 13, 2024, 10:12 p.m. • TechCrunch • (2 Minute Read)

OpenAI's recent partnerships with publishers, such as Le Monde and Prisa Media, to access French and Spanish news content for its ChatGPT chatbot have raised concerns about the impact on its competitors. The deals are part of OpenAI's efforts to expand its training data for AI models, with estimated payments ranging from $1 million to $20 million annually to publishers. This has sparked a debate about the barriers to entry for other AI vendors, as the high costs of licensing agreements could inhibit innovation and competition in the industry. Furthermore, questions about the ethical and regulatory implications of such deals and their potential impact on the news industry have been raised, underscoring the need for a nuanced approach to addressing these complex issues.

AI image-generator Midjourney blocks images of Biden and Trump as election looms

March 13, 2024, 9:57 p.m. • The Associated Press • (4 Minute Read)

The popular artificial intelligence image-generator Midjourney has started blocking its users from creating fake images of President Joe Biden and former President Donald Trump ahead of the upcoming U.S. presidential election. The CEO, David Holz, addressed the issue in a digital office hours event, stating the need to "put some foots down on election-related stuff for a bit." This moderation measure aims to deter potential abuse of the tool. The company's decision followed ongoing concerns about generative AI tools fueling election misinformation. The move may mitigate the risk of producing disinformation, as noted by the Center for Countering Digital Hate, which reported that Midjourney had been generating convincing images of political figures, potentially supporting disinformation or false election fraud claims. This action aligns with the tech industry's efforts to combat AI-generated deepfakes.

Watch: OpenAI powers Figure robot that can do human chores

March 13, 2024, 3:12 p.m. • VentureBeat • (3 Minute Read)

OpenAI has partnered with robotics startup Figure to debut a groundbreaking humanoid robot that showcases remarkable capabilities, including the ability to interact with humans and its environment. The Figure 01 robot, equipped with advanced neural networks provided by OpenAI, demonstrated its adeptness in following commands, identifying objects, engaging in conversation, and autonomously completing tasks such as picking up trash and putting away dishes. The robot's visual reasoning and language understanding, combined with its dexterous actions, exemplify a significant leap in general-purpose robotics interactions. Although this demonstration represents a prototype, Figure envisions training a world model to operate humanoid robots at a billion-unit level, with the goal of making a positive impact on humanity and alleviating the need for undesirable jobs. The collaboration between OpenAI and Figure is poised to raise the bar for humanoid robotics and exert pressure on competitors in the industry.

EU Approves World's First AI Regulations--Here's What To Know

March 13, 2024, 2:36 p.m. • Forbes • (2 Minute Read)

The European Union has approved the world's first regulations for artificial intelligence. The regulations aim to address concerns about the potential risks of rapidly developing AI technology. The AI Act places regulations on various AI technologies based on their potential risks and level of impact. It particularly focuses on high-risk AI systems used in critical infrastructure or medical devices, requiring them to assess and reduce risks, ensure transparency in data usage, and maintain human oversight. The regulations also ban certain AI applications that threaten citizens' rights, such as emotion recognition systems in schools and workplaces. The law is expected to be in effect by May, with a maximum fine of $38 million for violations. The development of regulatory rules governing chatbots is also set to be in effect by mid-2026. These regulations reflect efforts to ensure that AI technology remains human-centric, providing economic growth, societal progress, and unlocking human potential, as stated by Dragos Tudorache, a parliament member involved in drafting the AI Act. This move comes in the wake of calls from tech executives, including former Google CEO Eric Schmidt and OpenAI chief executive Sam Altman, for regulations to address the potential risks associated with AI.

I got a massage from an AI robot, and it was surprisingly relaxing

March 13, 2024, 2 p.m. • ZDNet • (6 Minute Read)

Aescape, a company specializing in AI technology, is set to introduce ten AI massage robot tables at Equinox fitness centers across New York City. The robot massage table, developed by Aescape's CEO Eric Litman, uses AI to scan the body and personalize the massage experience based on over 1 million 3D data points. It simulates a massage therapist's practice and allows for customized treatment plans, tailored to the individual's needs. The integration of technology in the massage therapy industry comes at a time of expected growth in the employment of massage therapists, offering a semi-autonomous and personalized massage experience for users. The 30-minute AI massage sessions will start at $60 and are expected to provide a new and accessible alternative to traditional massage therapy for many individuals.

Europe's AI Act: How does it work and what happens next?

March 13, 2024, 11:48 a.m. • The Associated Press • (5 Minute Read)

European Union lawmakers have given final approval to the 27-nation bloc’s artificial intelligence law, marking a significant milestone in the regulation of AI technology. The AI Act, which is expected to take effect later this year, is designed to prioritize human control over AI, allowing for economic growth and societal progress while safeguarding against potential risks. The law establishes a risk-based approach, with varying levels of scrutiny for different AI systems. It also addresses generative AI models and deepfakes, with provisions for transparency, data usage, and risk assessment. The EU's regulatory efforts are expected to influence global AI governance, as other countries and regions are also developing their own AI regulations. The EU will begin enforcing the AI Act in stages, with fines of up to 35 million euros or 7% of a company’s global revenue for violations. This law is just the beginning of the EU's AI-related legislation, as further regulations may be pursued in the future.

Perplexity brings Yelp data to its chatbot

March 12, 2024, 10:37 p.m. • The Verge • (2 Minute Read)

Perplexity, an AI chatbot, is integrating Yelp's data into its responses to provide users with multiple Yelp reviews, location information, and other details when seeking restaurant recommendations. Perplexity CEO Aravind Srinivas states that the integration aims to offer information directly from the source, catering to users' search for restaurants. The company will use Yelp's maps, reviews, and other data to enhance its responses, making it stand out compared to other chatbots. Yelp's SVP of business & corporate development, Chad Richard, expressed the company's willingness to provide Perplexity users with reliable local search results. This integration with Yelp marks Perplexity as the latest AI company to secure a source of up-to-date information, following similar moves by Google and OpenAI. Perplexity plans to pursue additional integrations in the future, potentially with providers of shopping and financial data.

Sam Altman's OpenAI comes out swinging against 'incoherent' and 'frivolous' Elon Musk in new lawsuit

March 12, 2024, 1:31 p.m. • Fortune • (3 Minute Read)

Sam Altman's OpenAI has entered into a legal battle against Elon Musk, with the non-profit filing a lawsuit in response to Musk's claims of violating the "Founding Agreement" and privatizing the development of the world's first artificial general intelligence, GPT-4. OpenAI has labeled Musk's lawsuit as "frivolous" and "incoherent," alleging that it is a ruse to gain access to their proprietary technology. In a filing with California's superior court, OpenAI's lawyers argued that Musk's claims are unfounded and an attempt to capitalize on their success without his involvement. OpenAI has also warned Musk that if he continues the lawsuit, more embarrassing documents could be revealed, asserting that evidence would show his prior support for a for-profit structure for OpenAI.

How OpenAI's text-to-video tool Sora could change science - and society

March 12, 2024, 12:30 p.m. • • (5 Minute Read)

OpenAI’s debut of its impressive text-to-video tool, Sora, has raised important questions regarding its potential impact on science and society. The California-based company showcased how Sora can create photorealistic videos from short text prompts, leading to concerns about the misuse of the technology. The tool's capabilities have sparked worries about the potential for misinformation and hateful content, particularly in the context of upcoming elections. However, experts also see the potential benefits of Sora, such as simplifying complex concepts for a lay audience and aiding in data management for scientific research. With such powerful technology at hand, society faces the challenge of developing new ways to evaluate the content produced. As these tools democratize the creation of media content, there is an urgent need to navigate the consequences of their widespread use.

How do neural networks learn? A mathematical formula explains how they detect relevant patterns

March 12, 2024, 11:23 a.m. • • (6 Minute Read)

A team of researchers led by data and computer scientists at the University of California San Diego has unveiled a mathematical formula that sheds light on how neural networks learn relevant patterns in data. This breakthrough provides a key to understanding the inner workings of neural networks, which have been powering advancements in artificial intelligence. The formula, based on a statistical analysis technique, offers a streamlined description of how neural networks learn features and make predictions. This insight not only helps in comprehending the decision-making mechanisms of neural networks but also paves the way for building simpler, more efficient, and more interpretable machine learning models. The study, published in the journal Science, marks a significant step toward developing a mathematical theory that explains how neural networks function and is part of a broader effort to bridge the gap between technology and theory in the field.

China puts trust in AI to maintain largest high-speed rail network on Earth

March 12, 2024, 10 a.m. • South China Morning Post • (3 Minute Read)

China is entrusting artificial intelligence to monitor and uphold the world's largest high-speed rail network, spanning 45,000km. The AI system, based in Beijing, processes real-time data and alerts maintenance teams of potential issues with 95% accuracy within 40 minutes. The technology has prevented speed reductions due to track irregularities and decreased minor track faults by 80% compared to the previous year. It also enhances infrastructure maintenance and safety. The adoption of AI and big data allows for precise assessments and scientific evaluations of infrastructure status. China's high-speed rail network, surpassing the length of the equator, operates at 350km/h, with plans to increase speed, presenting a maintenance challenge. The use of AI is improving maintenance efficiency and safety prevention, countering potential labor shortages, setting a new standard in rail maintenance.

Midjourney debuts feature for generating consistent characters across multiple gen AI images

March 12, 2024, 2:10 a.m. • VentureBeat • (4 Minute Read)

Midjourney, a popular AI image generating service, has recently introduced a new feature that allows users to recreate consistent characters across multiple generated images. This development addresses a key challenge faced by AI image generators, as these platforms often produce unique responses for each prompt entered into them. The new "–cref" tag enables users to specify character attributes and references, enhancing narrative continuity for visual storytelling in mediums such as films, novels, and graphic novels. Users can also control the level of similarity between new and original characters using the "–cw" tag, providing a degree of customization. This feature signifies a step towards Midjourney's evolution from a creative tool to a professional resource and is gaining traction among artists and creators. The feature is currently in the testing phase, with the official beta version approaching soon.

You can now make your own custom Copilot GPT. Here's how

March 11, 2024, 8:51 p.m. • ZDNet • (5 Minute Read)

In the era of AI chatbots, Microsoft has introduced a new feature for Copilot Pro users, allowing them to create their own custom Copilot GPT without any coding, even shareable with the Copilot community. To start, users can access the "Create a new Copilot GPT" option on the Copilot Pro homepage and follow the step-by-step instructions to build a chatbot for a specific task. Microsoft provides tips and prompts to guide users through the process and allows them to preview, tweak, and publish their custom GPTs. This feature aims to save time by personalizing chatbots for repetitive tasks, offering a more efficient AI tool for both personal and professional use.

Why Elon Musk Had to Open Source Grok, His Answer to ChatGPT

March 11, 2024, 6:15 p.m. • WIRED • (3 Minute Read)

Elon Musk has announced that he will release his AI chatbot, Grok, as an open-source platform following a lawsuit against OpenAI. Musk accused OpenAI of becoming too closed off and withholding technology, prompting him to make Grok available for anyone to download and use. By open-sourcing Grok, Musk aims to demonstrate his commitment to giving away artificial intelligence and to drum up interest in his company’s AI. This move aligns with Meta’s approach to generative AI and may help Grok become a powerful AI model in the future. However, it could also draw Musk into the ongoing debate over the benefits and risks of providing unrestricted access to advanced AI models.

Who are OpenAI's new board members?

March 11, 2024, 2:45 p.m. • Reuters • (1 Minute Read)

OpenAI, the company behind the ChatGPT, has announced its new board members, including returning CEO Sam Altman and three new directors. The board additions feature Sue Desmond-Hellman, former CEO of the Bill and Melinda Gates Foundation, Nicole Seligman, former president of Sony Entertainment, and Fidji Simo, CEO of Instacart and former head of Facebook. Simo brings extensive experience from the tech and social media industries, while Desmond-Hellman's background includes leadership roles at the Gates Foundation and other notable organizations, and Seligman's expertise lies in law and entertainment. These new appointments signal a significant shift for OpenAI as they work to move past recent controversies and continue their advancements in artificial intelligence technology.

Elon Musk says xAI will open source its Grok chatbot this week

March 11, 2024, 2:16 p.m. • Engadget

Elon Musk announced that xAI will be open sourcing its Grok chatbot, a decision set to take effect this week. This move follows Musk's recent lawsuit against OpenAI, a company he co-founded, for allegedly prioritizing profits over its original non-profit mission. Open sourcing Grok is seen as a strategic decision to increase the model’s uptake among third-party developers and researchers, and to gain valuable feedback for its improvement. Musk's earlier move to open source Tesla's patents, as well as X's publishing of its "For You" algorithmic feed code last year, reflects a broader trend in making technology more accessible to the wider community.