Agregátor RSS

Analytici ze SimilarWebu statisticky spočítali, která klíčová slova jsou na Googlu nejoblíbenější. Za předchozí rok, od března 2024 do února 2025, byl hlavním termínem youtube . Vyhrál už počtvrté za sebou, každý měsíc jej lidé do vyhledávače zadají v průměru půlmiliardkrát. Do top 10 se dostaly i ...

Non-tech companies trying out AI agents in the field said the technology still has a lot of  challenges to overcome before it can be used practically. 

Some of the early uses of agentic AI are in the area of customer service, which was part of a panel discussion at Nvidia’s GTC conference, held last week in San Jose.

The main problem is humans still prefer talking to humans because they don’t trust machines, said Cameron Davies, chief data officer at Yum Brands, which includes brands such as Pizza Hut, Taco Bell and KFC.

The company hopes to get to 100% digital ordering in five years, with agentic AI playing a role, Davies said.

Specifically, Yum Brands wants to automate orders at drive-throughs, Davies said. One of the most challenging service jobs is running drive-through windows — to take, fill and cash out orders — and AI agents could lighten that load. 

But the main issue isn’t technological, Davies said. “What’s the greatest challenge in putting this into place, is that nobody wants to talk to a machine right now. And you have to ask yourself why is that the case?” 

Yum Brands is also eyeing AI agents to reduce the “cognitive” load on human servers and employees. Scripted agents can do the role of “upselling and asking about charity donations,” Davies said.

“You do these things, then that person can now focus on being happier, making sure the order is right, getting the change right, etc.,” Davies said, adding that Yum Brands has been testing agents for back-office and HR functions, with mixed results.

Beyond trusting AI systems, compliance and accuracy are concerns for Craig Daniels, the head of Mayo Clinic’s Smart Hospital and Unbound Project.

Healthcare, by its nature, tends to lag other industries in adopting technology; Daniels is looking at the progress of AI agents at companies like Yum Brands to see what works.

For Mayo, the challenge is for doctors and patients to gain a high level of trust in AI to assist in diagnosis and treatment. Then they can consider the role of agents in helping doctors and patients.

Patients need to trust AI models just as they trust MRI machines without having to worry about the underlying technology, Daniels said.

Mayo Clinic is creating its own data platform with anonymous patient information gathered from 61 different healthcare organizations in four continents. For example, one Mayo AI system trained on 7 million electrocardiograms can detect and diagnose heart failure.

“There’s a point at which we have to trust the model works,” Daniels said. “It generates novel insights and we’ve researched and we trust that and we’re using it with the human to make the final decision. That’s a wonderful advancement.”

The US Food and Drug Administration (FDA) will ultimately regulate AI in healthcare through “medical-as-a-software device,” which will require proven research, Daniels said.

More than in other industries, Mayo will require a variety of guardrails be in place to deliver trustworthy results as there’s no allowance for AI or agents to hallucinate. Many panelists also mentioned they were still in the process of figuring out those guardrails. “We want to be safe,” Daniels said.

The panelists said that while customer service and chatbots aren’t new, they’re being revisited in light of the arrival of AI agents.

Agents are expected to grow more human-like in reaction and voice, and they’re more flexible. Agents don’t follow scripts like chatbots, and depending on the customer, they can connect to other agents to better serve customers.

“I can make it talk like a person. I can change dialects. I can do those things, and I want to control it,” Yum Brands’ Davies said.

Agents won’t fully replace people any time soon, as human ingenuity is still required in many areas — to supervise AI agents, modify them to be more effective, and be able to verify AI output and results.

U.S. Bank’s first AI agents are augmenting human knowledge to service customers.

Human agents for banking, mortgage or investments are subject matter experts that need information at their fingertips; this is where the bank is testing AI agents, said Sumitri Kolavennu, head of AI research and senior vice president at U.S. Bank.

“Keeping the human-in-the-loop in many [regulated] industries is … really paramount,” Kolavennu said. “We love the advantages and autonomy aspects of AI agents, but we do want to keep the human in the loop.”

U.S. Bank is testing the effectiveness of agents by seeing how quickly problems are resolved.

“The biggest thing we are seeing is resolution on first call. When you call, you don’t want the agent to say, ‘Let me call you back tomorrow’ or something like that. Those are some of the things that we are seeing [AI agents] being able to do,” Kolavennu said.

Mapy.cz rostou a postupně přechází na Mapy.com. V plánu je vylepšení Map novými zahraničními uživateli.

It's been a very busy week for Digicash Donald's administration

Analysis  Is the US retreating from its hardline stance on crypto? On Friday, the US Treasury Department lifted sanctions imposed on notorious crypto mixer Tornado Cash, once accused of washing billions in illicit crypto for criminals and nation-states alike.…

Tohle není jen velká powerbanka. Je dostatečně silná na to, aby utáhla bourací kladivo, silnou vrtačku, k tomu malou ledničku na svačinku a tekutiny na uhašení žízně. Navíc k ní snadno připojíte i „dospělé“ soláry.

A quiet tweak in a popular open-source tool opened the door to a supply chain breach—what started as a targeted attack quickly spiraled, exposing secrets across countless projects. That wasn’t the only stealth move. A new all-in-one malware is silently stealing passwords, crypto, and control—while hiding in plain sight. And over 300 Android apps joined the chaos, running ad fraud at scale behind

A quiet tweak in a popular open-source tool opened the door to a supply chain breach—what started as a targeted attack quickly spiraled, exposing secrets across countless projects. That wasn’t the only stealth move. A new all-in-one malware is silently stealing passwords, crypto, and control—while hiding in plain sight. And over 300 Android apps joined the chaos, running ad fraud at scale behindRavie Lakshmananhttp://www.blogger.com/profile/[email protected]

Byl představen Raspberry Pi PoE+ Injector pro napájení Raspberry Pi po datovém síťovém kabelu (PoE). Cena je 25 dolarů.

** Apple zvažuje, že do hodinek Apple Watch přidá fotoaparáty ** Mají mu pomocí s rozpoznáváním okolních objektů v rámci Apple Intelligence ** Samsung s foťáky v hodinách před lety neuspěl, co bude chtít Apple udělat jinak?

Cybersecurity researchers have uncovered two malicious extensions in the Visual Studio Code (VSCode) Marketplace that are designed to deploy ransomware that's under development to its users. The extensions, named "ahban.shiba" and "ahban.cychelloworld," have since been taken down by the marketplace maintainers. Both the extensions, per ReversingLabs, incorporate code that's designed to invoke a

Cybersecurity researchers have uncovered two malicious extensions in the Visual Studio Code (VSCode) Marketplace that are designed to deploy ransomware that's under development to its users. The extensions, named "ahban.shiba" and "ahban.cychelloworld," have since been taken down by the marketplace maintainers. Both the extensions, per ReversingLabs, incorporate code that's designed to invoke aRavie Lakshmananhttp://www.blogger.com/profile/[email protected]

If given the choice, most users are likely to favor a seamless experience over complex security measures, as they don’t prioritize strong password security. However, balancing security and usability doesn’t have to be a zero-sum game. By implementing the right best practices and tools, you can strike a balance between robust password security and a frictionless user experience (UX). This article

If given the choice, most users are likely to favor a seamless experience over complex security measures, as they don’t prioritize strong password security. However, balancing security and usability doesn’t have to be a zero-sum game. By implementing the right best practices and tools, you can strike a balance between robust password security and a frictionless user experience (UX). This articleRavie Lakshmananhttp://www.blogger.com/profile/[email protected]

Češi patří ke světové špičce ve vývoji protokolu Matter, který by jednou mohl konečně propojit všechny krabičky chytré domácnosti. Za projektem stojí největší hráči na scéně včetně Applu nebo Googlu a českou kartu drží společnost ELKO EP z moravského Holešova. [ArticleBox ORIGINÁL Co je to Matter ...

When generative AI (genAI) suddenly burst onto the tech scene with the arrival in late 2022 of OpenAI’s ChatGPT, companies quickly embraced its potential for automating tasks such answering customer inquiries, handling support tickets, and generating content.

A slew of rival chatbots followed ChatGPT. But they tended to be static tools; they didn’t learn from user interactions or application integrations. Only their foundational large language models (LLMs) could be trained.

Enter agentic AI. By leveraging technologies such as machine learning, natural language processing (NLP), and contextual understanding, AI agents can operate independently, even partnering with other agents to perform complex tasks.

“Think virtual coworkers able to complete complex workflows,” McKinsey & Co. explained in a report. “The technology promises a new wave of productivity and innovation.”

According to the 2025 Connectivity Benchmark Report from Mulesoft and Deloitte, 93% of IT leaders plan to introduce autonomous AI agents within two years — and nearly half have already implemented them.

Like chatbots, AI agents have existed since the 1960s. However, it wasn’t until advances in AI, ML, deep learning, and transformer models (such as GPT-3 and ChatGPT) that they became capable of adapting to tasks and learning from data. That dramatically expanded their use cases.

Agentic AI systems typically use a transformer-based LLM as the core, enhanced with reasoning, memory, reinforcement learning, and tool integrations. The LLM’s understanding of language allows it to interpret instructions and generate responses.

In the simplest terms, an AI agent is the combination of an LLM and a traditional software application that can act independently to complete a task. They can operate autonomously, make decisions, plan, and take actions to achieve specific goals without constant human oversight.

“This is a way to deliver business value, and I think that is where the focus should be, to think about how you’re going to disrupt the business process,” said Samta Kapoor, a principal on Ernst & Young’s tech consulting team.

For example, if an employee requests vacation time, an AI agent can automate the process of entering the dates into the HR system and ensuring all other systems are aware that employee will be away for the specified time. If the employee changes his or her mind and enters new dates, the agent can reschedule everything in the HR system autonomously. All it takes is a simple set of commands and away the AI agent goes, Kapoor said.

AI agents can also autonomously write software code and offer that base code to a developer, who can then review it for accuracy and modify it if necessary. But there are also agents that can perform the code review, as well. And, best of all, it can all be done in seconds, not hours or days.

AI-assisted code generation tools are increasingly prevalent in software engineering and, somewhat unexpectedly, have become low-hanging fruit for most organizations experimenting with generative AI (genAI) tools. Adoption rates are skyrocketing, because even if they only suggest a baseline of code for a new application, automation tools can eliminate hours that otherwise would have been devoted to manual code creation and updates.

By 2027, 70% of professional developers are expected to be using AI-powered coding tools, up from less than 10% in September 2023, according to Gartner Research. And within three years, 80% of enterprises will have integrated AI-augmented testing tools into their software engineering tool chain — a significant increase from approximately 15% early last year, Gartner said.

Beyond coding, AI agents are designed to perceive their surroundings, make decisions based on that information, take actions, and sometimes learn and adapt over time to perform tasks autonomously. Reinforcement learning is key to agentic AI’s ability to continue to grow in sophistication when performing tasks.

“If you’re playing a game, you either win or lose. If you lose, you go back and evaluate why, and then play again but do it differently,” Kapoor said. “With agentic AI, there are a very defined set of KPIs that you’re asking it to meet, so it would know whether it has met them or not. And then it goes back and it reinforces itself to do this task differently.”

For agentic AI, decision-making is structured around autonomy and goal-orientation. “There is a reward system within agentic AI and this is frequently based on reinforcement learning, where the AI learns to maximize rewards through interactions with its environment,” said Arun Gururajan, NetApp’s vice president of research and data science.

The sense-think-act process and agent types

Agentic AI, Gururajan said, follows a cyclical sense-think-act process, which has the following steps:

1. Perception: The system gathers input from its environment and/or the user.
2. Reasoning and Planning: The central brain of the agent, typically a powerful LLM, reasons through the task and generates and evaluates possible actions.
3. Decision-making: Reinforcement learning strategies, often supplemented by human feedback as well as the memory of past interactions, help select the optimal action.
4. Execution: The chosen action is carried out, possibly by calling on internal/external tools via API integrations.
5. Feedback loop: Outcomes are assessed and used to refine future decisions, creating a continual learning process.

There are several types of AI agents that can be employed based on the complexity of the task. They include:

• Reactive agents: These only respond to their environment based on predefined rules. They don’t store history or learn from it (e.g., simple game AI). The most basic of agents, they’re used in customer service bots or smart home devices that can adjust themselves automatically.
• Deliberative agents: These use an internal model and reasoning to make informed, long-term decisions. They’re used in applications such as autonomous vehicles, supply chain management, and medical decision systems.
• Hybrid agents: These combine reactive and deliberative approaches for more efficient decision-making. For instance, a robot might react to immediate obstacles and plan its path to a goal simultaneously. Hybrid AI is used in automating business tasks, where reactive agents handle routine actions (e.g., responding to emails) while deliberative agents plan and optimize workflows for efficiency over time.

In short, hybrid agents integrate both immediate reaction and thoughtful planning in their decision-making.

“Traditional AI — or predictive AI — is often tuned to solve a narrow and specific problem — for example, predicting drive failure in storage systems,” Gururajan said. “Agentic AI is more dynamic; It can adapt, reason and strategize.”

Imagine, for example, agents managing a data storage system by monitoring dashboards, identifying bottlenecks, predicting failures, and proactively taking action to prevent errors, ensuring system SLAs are met.

NetApp, for instance, sets up reward models based on objectives (such as maximizing uptime or minimizing energy use) that combine human preferences, real-time data, and instructions, enabling AI to optimize behavior and improve performance over time, according to Gururajan.

Reasoning techniques like Chain-of-Thought prompting, which mirrors human thought, or ReAct prompting help break down tasks and plan actions. Memory modules store context and intermediate results for tasks requiring continuity. Reinforcement learning with human feedback fine-tunes the system’s outputs to align with human values. Additionally, tool integrations enable the AI to perform complex tasks beyond text generation, such as web search and interacting with APIs.

The growing use of API integrations

API integrations with AI agents are currently the pinnacle of use cases. In agentic AI, tools via API integrations allow agents to interact with the real world. When a task requires external information, the agent generates an API call, formats parameters, authenticates, and processes the response to complete the task or take further action.

“When an agent needs to perform a task that requires external information, such as searching a database, sending an email, executing another ML model,”  Gururajan said, “it generates an API call based on its understanding of the task and the API’s documentation.

Executing on an API involves formatting aspects with the correct parameters and authenticating with the API, which in turn, returns data (or performs actions); the agent processes the response and completes the task or takes subsequent actions if needed, Gururajan explained.

Looking ahead, there are still improvements needed for agentic AI to mature, such as addressing challenges with API discoverability and adaptation, and dealing with issues such as a lack of standardization and documentation, Gururajan said.

Change management also makes it difficult for agents to select the right APIs. And API security and authentication remain crucial, requiring robust protocols and access control to protect sensitive data. Implementing service-level credentials could provide more granular control, such as restricting agents to read-only access or specific actions.

There is emerging research involving agents, such as multi-objective optimization, which focuses on solving conflicting task goals using goal-based programming. Additionally, system-level heuristics can be created as general rules reflecting core principles, constraints, or safety measures.

Heuristics can be incorporated into the agentic framework by: (a) filtering goals (such as removing goals requiring restricted data), (b) modifying objectives (insuring safety overrides efficiency), and (c) integrating reinforcement learning to weight goals.

Looking ahead, there’s a need for agents to autonomously create their own APIs for tasks, as most agents currently rely on pre-existing ones. “This would be a positive step towards Artificial General Intelligence or AGI,” Gururajan said.

 

Throw a spanner in the works, best get good at fixing things. Now, where did you put that spanner?

Opinion  Never attribute to malice that which is adequately explained by stupidity. This works well in sane times, less so when "but it's both" is the default. Apply it to Microsoft's decision to make bug reports include not only a working example but a video of the same, and the meter oscillates wildly. What were they thinking? What did they expect?…

A critical security flaw has been disclosed in the Next.js React framework that could be potentially exploited to bypass authorization checks under certain conditions. The vulnerability, tracked as CVE-2025-29927, carries a CVSS score of 9.1 out of 10.0. "Next.js uses an internal header x-middleware-subrequest to prevent recursive requests from triggering infinite loops," Next.js said in an

A critical security flaw has been disclosed in the Next.js React framework that could be potentially exploited to bypass authorization checks under certain conditions. The vulnerability, tracked as CVE-2025-29927, carries a CVSS score of 9.1 out of 10.0. "Next.js uses an internal header x-middleware-subrequest to prevent recursive requests from triggering infinite loops," Next.js said in an Ravie Lakshmananhttp://www.blogger.com/profile/[email protected]