Monthly Archives: May 2024

Metaschief AI scientist thinks large language models will never reach human intelligence.

Yann LeCunasserts that artificial intelligence (AI) large language models (LLMs) such as ChatGPT have alimited grasp on logic, the Financial Times (FT) reported Wednesday (May 21).

These models, LeCun told the FT, do not understand the physical world, do not have persistent memory, cannot reason in any reasonable definition of the term and cannot plan...hierarchically.

He argued against depending on LLMs to reach human-level intelligence, as these models need the right training data to answer prompts correctly, thus making them intrinsically unsafe.

LeCun is instead working on a totally new cohort of AI systems that aim to power machines with human-level intelligence, though this could take 10 years to achieve.

The report notes that this is a potentially risky gamble, as many investors are hoping for quick returns on their AI investments. Meta recently saw its value shrink by almost $200 billion after CEO Mark Zuckerbergpledged to up spendingand turn the tech giant into the leading AI company in the world.

Meanwhile, other companies are moving forward with enhanced LLMs in hopes of creating artificial general intelligence (AGI), or machines whose cognition surpasses humans.

For example, this week saw AI firmScaleraise $1 billion in a Series F funding round that valued the startupat close to $14 billion, with founder Alexandr Wang discussing the companys AGI ambitions in the announcement.

Hours later, the French startup called H revealed it had raised $220 million, with CEO Charles Kantor telling Bloomberg News the company is working towardfull-AGI.

However, some experts question AIs ability to think like humans. Among them isAkli Adjaoute, who has spent 30 years in the AI field and recently authored the book Inside AI.

Rather than speculating about whether the technology willthink and reason, he views AIs role as an effective tool, stressing the importance of understanding AIs roots in data and its limitations in replicating human intelligence.

AI does not have theability to understandthe way that humans understand, Adjaoute told PYMNTS CEO Karen Webster.

It follows patterns. As humans, we look for patterns. For example, when I recognize the number 8, I dont see two circles. I see one. I dont need any extra power or cognition. Thats what AI is based on. Its the recognition of algorithms and thats why theyre designed for specific tasks.

Read the rest here:

Meta AI Head: ChatGPT Will Never Reach Human Intelligence - PYMNTS.com

The House Homeland Security Committee may host town hall meetings for lawmakers on artificial intelligence and could bring in tech experts to help lawmakers better understand rapidly evolving technology before writing rules to govern it.

During a Wednesday hearing focused on how AI can be used to secure and defend the U.S., lawmakers acknowledged a new format may help them get their arms around a deeply complex set of issues.

What I may do is have a town hall type thing, where were the town hall and theyre on the [dais] and were just asking questions, Rep. Mark Green, Homeland Security Committee chairman and Tennessee Republican, said during Wednesdays hearing. I think that would be more informative. And maybe some presentations, so to speak, on data poisoning for AI and all that kind of stuff.

The town hall format would be familiar to tech experts who frequently field requests from Capitol Hill. Senate Majority Leader Charles E. Schumer convened AI insight forums starting last year. The New York Democrat also worked with a group of three other lawmakers to develop comprehensive AI legislation.

Mr. Schumers closed-door forums attracted top tech minds such as Elon Musk, Bill Gates, Metas Mark Zuckerberg and Googles Sundar Pichai, and theevents gave senators a chance to ask questions without the usual time constraints of formal Capitol Hill hearings.

But the forums also had their share of detractors.

SEE ALSO: Seoul AI summit aims to fill regulatory vacuum, but critics say voluntary pledges fall short

Sen. Josh Hawley, Missouri Republican, criticized the forums as a giant cocktail party for Big Tech. At the same time, Democratic-led committees in the Senate did not appear thrilled about the potential for Mr. Schumers effort to encroach on legislation under their jurisdiction.

After the Senate events, Mr. Schumers group last week produced an AI roadmap. It didnot propose specific legislation, but did call for at least $32 billion more spending on tech research and development.

The House Homeland Security Committees purview is more narrow and fixated on threats involving cybersecurity, infrastructure, and Americas physical borders, among other things.

In the absence of congressional rules, President Biden is implementing an AI executive order that includes guidance for federal agencies.

Mr. Biden has also dispatched administration officials to meet with foreign governments on AI rules, including a first meeting with Chinas government in Geneva last week.

Europe is pressing ahead with new AI rules. In March, the European Parliament approved the EU AI Act, which bans various AI applications such as emotion recognition and scraping of facial images.

EU states gave additional agreement to the law this week, which the Council of the EU portrayed on Tuesday as the final green light on the new AI rules.

View original post here:

Artificial intelligence town halls? House committee weighs new approach before writing AI rules - Washington Times

Someday and it may be soon conversations withAmazonsautomated voice assistantAlexamay cost money. Amazon has been clear about its intentions to charge a subscription for a tiered Alexa service since last September.Butimprovements to voice technology driven by artificial intelligence (AI) may force the issue.

Asan Amazon spokesperson confirmed to PYMNTS Wednesday (May 22), that integrating generative AI into Alexa is in development,and was mentioned by CEO Andy Jassy in his most recentshareholder letterin which he referenced AI as part ofdozensof potential applicationsincludinganeven more intelligent and capable Alexa.Other company executives past and present have been on the record about the improvements AI can bring to the technology as well asthe costs it will add to production.

But before we would start charging customers for this and I believe we will it has to be remarkable,David Limp, the former Amazon executive in charge ofAlexa,toldBloombergin September 2023.It has to prove the utility thatyourecoming to expect from thesuperhumanassistant. I can paint a line from whereweregoing to be right now, whichwerenot going to charge for, to something that will have so much utility for everymember of the household. Wedonthave an idea of a price yet.Welltalk to customers and learn fromthem,what they believe the value is.

When will that happen? Morerecent reportsput the debut of Alexa Plus by the end of the year, but there is no more specific information. The only Alexa application currently under a subscription model isAlexa Together,whichwasdevelopedfor elderly care and connectivity among family caregivers. However, the company recently announced thatit will not be supportedafter June 25.

Using AI to power voice commerce and intelligent conversations was a development predicted by PYMNTS CEO Karen Webster as early as April 2023. In a column she wrote at that time, she predicted that technology would integrate voice and enable commerce from any connected device, positioning voiceto bea dominant means of interactions and transactions.Andshenoted,researchfrom PYMNTS Intelligencethat showed nearly 30% of U.S. consumers at that time would pay a monthly fee to access a voice assistant thatcanhandle complex commands and commerce.Not to mention her predictions about AI.

Artificial intelligencewill make voice interactions smart, personalized, adaptiveandengaging,she wrote.As in truly engaging conversational in every sense of the word. Not just reactive to a wake word and a series of prompts, but proactive and intuitive, anticipating actions based on history and context and anticipating what consumers might want to do next, just like an effective, smart, capable human assistant would do.

According to a September 2023 blogpost,the new Alexa might have been listening. It willbe basedon a new large language model (LLM)thatsbeen custom-built for voice interactions and aimed at getting real-time information, efficient smart home control, and maximizing home entertainment. The postsets outfive foundational capabilities: conversation, real-world utilities like smart home commands, personalization, personalityandtrust.

The emergence of AI, along with sophisticated chatbots and voice functionalities, is transforming the landscape of digital assistants at a time when consumers increasingly rely on technology to manage various daily tasks.

According toPYMNTS Intelligence, consumers devote 26% of their weekdaytimeand 28% of their weekend time to multitasking.While one might typically associate the integration of digital technology with activities like online shopping which accounts for only about a quarter of multitasking time the rest is spentmanaging work-related tasks, home upkeep, and staying in touch with family members, among others.

In the April 2023 study titled How Consumers Want to Live in the Voice Economy,PYMNTS Intelligence discovered that consumers are eager to streamline their daily routines usingsmart, simple and more interconnected solutions. Hands-free voice technologies are increasingly popular for tasks such asretrieving information, voice identification, or booking flights. On average, consumers engaged in six different tasks using voice technology over the past year.

The data also highlights a growing preference for voice-driven interfaces, with 54% of surveyed consumers indicating they would choose voice technology in the future for its speed over traditional typing or touch interactions.

For all PYMNTS AI coverage, subscribe to the daily AINewsletter.

More here:

Waiting for Alexa: When Will Amazon Strike Up the AI Conversation? - PYMNTS.com

This report finds that AI is diffusing broadly across technologies, inventor-patentees, companies, and U.S. geography.

October 2020 Read IP Data Highlight 5 Supplementary material

Technology diffusion is the spread and adoption of a new technology by inventors, companies, and other innovators. Technologies that diffuse broadly have potentially large effects on innovation, productivity, and economic growth. For example, steam power, electricity, and information technology greatly enhanced the volume, as well as the variety, of goods produced within the economy.

The figure below shows that the percentage of U.S. organizations (green line) and inventors (dashed blue line) that patent in AI increased from under 5% in 1980, to just over 20% in 2018. This is remarkable growth and shows that AI is increasingly important to U.S. invention.

Image

Follow this link:

Inventing AI: Tracing the diffusion of artificial intelligence with U.S. patents - United States Patent and Trademark Office

With this latest agreement, BBVA is once again ahead of the curve when it comes to embracing disruptive technologies that will impact the financial industry. Notably, it is the first European bank to forge an alliance with OpenAI, which will share its knowledge and unlock the full potential of the new tool at the bank.

In the past two years, BBVA added 7,187 professionals in the data and technology field to its workforce - a figure it plans to increase in 2024 with 2,700 new hires. Of this amount, 1,225 will take place in Spain for the banks headquarters in Madrid, Bilbao and Barcelona. Technology talent has become a key aspect of the Groups digital strategy.

BBVA has already begun deploying 3,000 ChatGPT Enterprise licenses among Group employees in a bid to increase productivity and process efficiency, while stimulating innovation across the Group. The enterprise version of ChatGPT delivers the utmost security and privacy, combined with its unique ability to generate content or answer complex business questions, among numerous other features.

OpenAI has also agreed to deliver training and provide the latest updates for its large language models (LLMs), the technology on which ChatGPT is built. By working in close partnership with OpenAI, BBVA will drive forward the most successful use cases for the banks business and processes.

Data and technology are the key levers of transformation at BBVA, which for over a decade has been running specific development centers for advanced analytics and artificial intelligence, now known as AI Factories, in Spain, Mexico and Trkiye. In addition to these teams, there are also analysts and data specialists working across all the business areas, giving a total of more than 5,000 employees, of whom 1,000 or so are data scientists. The entire team works to unlock the value of data and AI with a view to improving business decision-making and helping create various kinds of products, such as solutions to improve the financial health of customers or aid them in their climate transition.

The recent agreement with OpenAI is a further example of BBVAs ongoing commitment to generative AI as a key differentiating aspect in the value proposition it offers its customers. New artificial intelligence tools are going to have a disruptive impact on society as a whole and on the financial industry in particular. At BBVA, we want to further promote our role as pioneers when it comes to innovating in financial services and we are therefore firmly committed to exploring the potential of this technology. We believe that generative AI, when used safely and responsibly, is a game-changer in how we support our customers in their decisions and offer personalized services. It also happens to stimulate creativity among our employees, explains Ricardo Martn Manjn, Global Head of Data at BBVA.

OpenAI COO Brad Lightcap added: "Were excited to partner with BBVA, one of Europe's top banks, to deliver the capabilities of ChatGPT Enterprise at scale. AI streamlines mundane tasks, boosting creativity, efficiency, and productivity.

The bank is already handing out licenses at its central services in Spain, and this process will continue in the Groups other main countries. Compulsory training will be delivered for each account set up. The aim is for all areas and departments to have access to ChatGPT, so that licensed employees can collaborate with their colleagues in undertaking various projects. In tandem, BBVA will be collecting feedback and suggestions from these users through a multi-country community, with the aim of flagging the most outstanding use cases and sharing best practices.

We see this first foray into the use of ChatGPT Enterprise as an opportunity to validate the extent to which these tools can genuinely boost our productivity, thus transforming the way we all work within the bank, explains Elena Alfaro, the new Head of AI Adoption at BBVA, whose role is to mainstream this technology across the Group. We are aiming to enhance the capabilities of our employees, not to replace them, she says.

Alfaro also remarks that while ChatGPT Enterprise is certainly a major strategic commitment, it will not be the only solution to be used within the organization. BBVA is continuing to evaluate other tools that may prove viable for the more than 100 use cases to be rolled out over the course of 2024.

View original post here:

BBVA steps up its plans in artificial intelligence by signing an agreement with OpenAI - BBVA

MORGANTOWN, W.Va. Researchers from West Virginia University, Marshall University, and Florida International University are exploring the cybersecurity needs of artificial intelligence technologies with a $1.75 million grant from the Defense Advanced Research Projects Agency (DARPA).

Professor and Chairman of the Statler College of Engineering and Mineral Resources in the Lane Department of Computer Science and Electrical Engineering, Anurag Srivastava, said the AI-CRAFT project is intended to develop ways to secure the emerging technology. Artificial intelligence is developing rapidly and being pushed into larger real-time applications.

Our goal is to look at what those are, and how do I defend myself if someone is trying to hack into or make the AI behave in a way it is not supposed to behave? Srivastava said.

The teams are building the artificial intelligence systems while engineering security and safety as they are deployed. The complexity results from the evolution from simple calculation systems used a decade ago to the addition of millions of data points as AI systems are taught to think like the human brain.

Look at this from a new point of view now; what is the attack vector now? Srivastava asked. Can someone reverse engineer the AI? Can someone poison the AI so it behaves in a way it should not?

AI technologies have quickly grown from application in autonomous cars to use in very lifelike robots and vital systems like public utilities. The research also includes developing secure data practices, access controls, and continuous monitoring to assess the security and usefulness of AI systems.

Especially those that will be used to operate critical systems like robots or the power grid, Srivastava said.

On the academic side, students will have many hands-on opportunities in labs and training platforms, designed to equip students with the skills and knowledge needed to thrive in a rapidly evolving cybersecurity industry.

Other than solving this problem for complex systems like a power grid, robotics, or autonomous cars, our goal is also to teach it because this is also a new topic, Srivastava said.

Officials from WVU, Marshall University, and the U.S. Department of Defense will break ground on the new Institute for Cyber Security in Huntington on May 17.

Link:

Study including WVU and Marshall analyzes cyber threats to Artificial Intelligence systems - West Virginia MetroNews