AI trends 2025

830 Comments

1. February 20, 2025 at 8:39 am

   Tomi Engdahl says:

   Ben Schoon / 9to5Google:
   Google is removing Gemini support from the main Google app on iOS, pushing users to download the standalone Gemini app, which it launched in November 2024 — Google is informing iOS users that it will be removing Gemini support from the main Google app as it pushes users over to the full Gemini app.

   Google app on iOS removing Gemini as it pushes users to full app
   https://9to5google.com/2025/02/18/google-app-removes-gemini/

   Google is informing iOS users that it will be removing Gemini support from the main Google app as it pushes users over to the full Gemini app.

   One of the first ways to access Gemini on iOS was through the main Google app, which offered a switcher between Google Search and Gemini for several months. However, as Google has built out the Gemini experience, new features have been missing from the Google app – a key example being Gemini Live. That’s because Google has been focused on the full Gemini app for iOS, which launched in November.

   Reply
2. February 20, 2025 at 8:41 am

   Tomi Engdahl says:

   Adamya Sharma / Android Authority:
   Google is bringing Circle to Search-like visual searches to its Chrome and Google iOS apps, without the Circle to Search branding, as a Google Lens feature — The feature will work within the Chrome and Google iOS apps via Lens. — • — • — • — TL;DR

   Google rolls out visual search on iPhones, but curiously ditches ‘Circle to Search’ branding
   The feature will work within the Chrome and Google iOS apps via Lens.
   https://www.androidauthority.com/google-brings-circle-to-search-to-iphones-3527608/

   Circle to Search is one of Google’s most prominent and promoted AI features right now. Initially debuting on Samsung phones, its development is something Google manages, including updates and expansion to new devices. Now, Google is bringing Circle to Search to iPhones, but with a key difference: while it’s called “Circle to Search” on Android, Apple users will not experience the feature under the same name.

   The functionality of the new Lens feature on iPhones is almost identical to how Circle to Search works on Android. Whether browsing articles, shopping, or watching videos, iPhone users can now perform a visual search seamlessly while staying within their current page, without the need to take a screenshot or open a new tab.

   Reply
3. February 20, 2025 at 8:41 am

   Tomi Engdahl says:

   Jon Keegan / Sherwood News:
   Perplexity open sources R1 1776, a version of the DeepSeek R1 model that Perplexity CEO Aravind Srinivas says was “post-trained to remove the China censorship” — But shortly after its release, attention turned to how compliant the model was with Chinese censorship laws.

   Perplexity claims to have purged Chinese censorship and propaganda from its new DeepSeek clone
   When DeepSeek R1 was released, it shocked the AI world.
   https://sherwood.news/tech/perplexity-claims-to-have-purged-chinese-censorship-and-propaganda-from-its/

   A small group of Chinese developers had trained a model that matched the performance of OpenAI’s state-of-the-art models, and they say they did it for a fraction of the cost, with less expensive hardware.
   Open-sourcing R1 1776
   Open-sourcing R1 1776

   But shortly after its release, attention turned to how compliant the model was with Chinese censorship laws.

   Much like Meta’sMETA $700.20 (-1.76%) Llama 3 model, DeepSeek R1 model was released as open-source software, anyone could take the model and post-train, distill, or change it for any application. That’s exactly what AI startup Perplexity did.

   Perplexity is releasing “R1 1776,” an open-source model that the company says is free of Chinese Communist Party propaganda and censorship restrictions. Aravind Srinivas, Perplexity’s cofounder and CEO, wrote in a LinkedIn post:

   “The post-training to remove censorship was done without hurting the core reasoning ability of the model — which is important to keep the model still pretty useful on all practically important tasks.

   Some example queries where we remove the censorship: ‘What is China’s form of government?’, ‘Who is Xi Jinping?’, ‘how Taiwan’s independence might impact Nvidia’s stock price’.”

   Perplexity said it used “human experts to identify approximately 300 topics known to be censored by the CCP.”

   Reply
4. February 20, 2025 at 8:43 am

   Tomi Engdahl says:

   Melissa Heikkilä / Financial Times:
   Google launches a Gemini 2.0-based AI co-scientist tool to help biomedical scientists create novel hypotheses and speed up research, available to select testers

   Google builds AI ‘co-scientist’ tool to speed up research
   Lab assistant powered by artificial intelligence can help generate scientific hypotheses
   https://www.ft.com/content/6e53cc55-9031-4ba4-9e7c-e5e9c02b3203

   Google has built an artificial intelligence laboratory assistant to help scientists accelerate biomedical research, as companies race to create specialised applications from the cutting-edge technology.

   The US tech group’s so-called co-scientist tool helps researchers identify gaps in their knowledge and propose new ideas that could speed up scientific discovery.

   “What we’re trying to do with our project is see whether technology like the AI co-scientist can give these researchers superpowers,” said Alan Karthikesalingam, a senior staff clinician scientist at Google.

   Google’s new tool comes as tech companies are spending billions of dollars on AI models and products, believing the technology can change industries from healthcare to energy and education.

   Reply
5. February 20, 2025 at 8:55 am

   Tomi Engdahl says:

   Why the Network Matters to Generative AI
   https://www.networkcomputing.com/network-management/why-the-network-matters-to-generative-ai

   Multi-cloud networking is becoming increasingly important to the success of generative AI because the architectural pattern—whether at the board or application layer—always depends on the ability to transfer data between components.

   You know, where buses connect components like CPU, ALU, RAM, and, of late, GPU and DPU. Design of these systems requires answering questions about how fast the bus speeds between components must be and how much bandwidth is required to support a given set of performance requirements. This is where technologies like I2C, PCI, and QPI fit, why FSB is no longer used, and why DDR replaced SDR. The “network” that connects circuit-level components is a significant factor in processing speed and capacity.

   Applications, today, are distributed. Our core research tells us more than half (60%) of organizations operate hybrid applications; that is, with components deployed in core, cloud, and edge locations. That makes the Internet their network, and the lifeline upon which they depend for speed and, ultimately, security.

   Furthermore, our focused research tells us that organizations are already multi-model, on average deploying 2.9 models. And where are those models going? Just over one-third (35%) are deploying in both public cloud and on-premises.

   Applications that use those models, of course, are being distributed in both environments. According to Red Hat, some of those models are being used to facilitate the modernization of legacy applications. Legacy apps are typically on-premises, even if the AI used to modernize is it somewhere else.

   The role of multi-cloud networking

   So, we’ve got applications and AI distributed across the Internet, and a network that needs to connect them. Oh, and it’s got to be secure as well as fast.

   This is why we’re seeing so much activity focused on multi-cloud networking solutions. The misnamed technology trend (it’s not just about multiple clouds but about interconnecting multiple locations) is a focus on the network and a recognition of the important role it plays in securing and delivering applications today.

   For one thing, over-the-Internet connectivity doesn’t typically reach into another environment, in which there are all kinds of network challenges like overlapping IP addresses, not to mention the difficulty in standardizing security policies and monitoring network activity.

   These are the problems multi-cloud networking solves for. Multi-cloud networking basically extends a network into multiple environments rather than just connecting those environments via two secure endpoints, a la a VPN.

   Multi-cloud networking is becoming increasingly important to the success of generative AI because the architectural pattern—whether at the board or application layer—always depends on the ability to transfer data between components safely, reliably, and as fast as possible. Multi-cloud networking introduces some of the control network professionals are missing when they have to use the Internet as their network.

   Reply
6. February 20, 2025 at 9:02 am

   Tomi Engdahl says:

   A Quick Look into ChatGPT’s Network Traffic
   https://www.keysight.com/blogs/en/tech/nwvs/2023/02/21/a-quick-look-into-chatgpts-network-traffic

   Reply
7. February 20, 2025 at 11:18 am

   Tomi Engdahl says:

   Agentic AI: How It Works, Benefits, Comparison With Traditional AI
   Learn about agentic AI, how it works, its applications and challenges, and how it differs from traditional AI.
   https://www.datacamp.com/blog/agentic-ai
   AI is moving beyond just responding to prompts with text or images. The next big step is agentic AI, which gives AI systems the ability to understand their environment, set goals, and take action to achieve them.

   According to a recent report by Gartner, less than 1% of enterprise software applications used agentic AI techniques in 2024. According to the report, that number could rise to 33% by 2028. This means that more and more businesses are starting to see great value in agentic AI and hopefully after reading this article you will see why.

   What Is Agentic AI?

   There is currently no universally agreed-upon definition of agentic AI. Some people use the terms “agentic AI” and “AI agents” interchangeably, while others draw distinctions between them. Despite these differences, most agree on certain aspects of agentic AI.

   Agentic AI refers to AI systems that can operate with a degree of independence, making decisions and taking actions to achieve specific goals. Unlike traditional AI, which requires explicit prompts to generate results, agentic AI can analyze situations, develop strategies, and execute tasks in parallel. Agentic AI applications maintain control of how they accomplish tasks by using tools and making decisions about internal processes.

   Some key aspects that define agentic AI include:

   Autonomy: The ability to function without continuous human interaction.
   Goal-oriented behavior: Setting and pursuing objectives based on predefined or evolving goals.
   Adaptability: Responding to changing environments and learning from past interactions.
   Interoperability: Ability to use different data sources, tools, and platforms to enhance decision-making.

   To better understand what agentic AI is, think of an autonomous vehicle that continuously recalculates the optimal route (for speed or economy) to its destination as conditions change. This demonstrates autonomy in decision-making, goal-oriented behavior by maintaining focus on reaching the destination, adaptability to changing road conditions, and interoperability with existing infrastructure like roads, traffic signals, and real-time traffic data.

   Agentic AI vs. Traditional AI

   The difference between traditional AI and agentic AI comes down to autonomy and adaptability.

   Traditional AI is designed for specific tasks based on predefined rules or training data. These models analyze input and return outputs but don’t make independent decisions beyond their programming. Examples include recommendation systems, chatbots, and predictive models.

   Agentic AI goes beyond passive responses by actively planning, adapting, and making decisions in real time. It can interact with multiple systems, use external tools, and even refine its own objectives. Instead of waiting for user input, it can initiate tasks, learn from feedback, and self-improve.

   How Does Agentic AI Work?

   Agentic AI isn’t a single technology but a way of designing AI systems that operate with more independence than traditional models. The details vary from one application to another, but most agentic AI systems involve multiple LLMs that communicate through prompts, use external tools, and can read and write files. These systems often work asynchronously, making them feel more like distributed networks than isolated models.

   Reply
8. February 20, 2025 at 11:18 am

   Tomi Engdahl says:

   https://www.linkedin.com/pulse/industrial-iot-ai-generative-agentic-actual-case-studies-bottacci-kcbef

   Reply
9. February 20, 2025 at 11:20 am

   Tomi Engdahl says:

   Automation, Autonomy and Accountability, Agentic AI in 2025
   https://www.iotworldtoday.com/artificial-intelligence/automation-autonomy-and-accountability-agentic-ai-in-2025

   Experts say Agentic AI will revolutionize industries in 2025, enhancing autonomy, accountability and automation across sectors

   Agentic AI, or AI agents, are AI systems that can act autonomously to achieve goals with minimal human intervention. They independently take actions, adapt in real time and solve multi-step problems based on context and objectives to make decisions, plan, learn and communicate.

   Agentic AI went mainstream in 2024 and is expected to surge in 2025. IoT World Today has collected agentic AI predictions from companies across industries, examining how they expect agentic AI to transform industries in the coming year.

   Here are some of the major trends experts anticipate for 2025, including operational transformation, workflow automation, accountability and tool downsizing.

   Jill Goldstein, global managing partner of HR and talent transformation at IBM Consulting:

   “We’re entering a new chapter in how employees get work done with the rise of AI agents. Unlike AI assistants, AI agents can generate plans based on a prompt and carry out tasks independently. They are most effective when focused on specialized tasks and working together with other agents on complex, multipart requests. As AI agents become more common, companies will need to reevaluate their work processes and create new types of teams where humans oversee groups of autonomous AI agents.”

   Charles Crouchman, chief product officer at Redwood Software:

   “Agentic AI will push automation solutions beyond simple task execution, enabling intelligent decision-making on which tasks to run, when and in what order based on context — not just pre-set schedules or triggers.”

   Nitesh Bansal, R Systems CEO:

   “In the coming years, generative AI copilots, equipped with expanded memory and an agentic mesh framework, will increasingly enhance the software development life cycle (SDLC). As these models gain the capability to process large codebases, integrate complex documentation and third-party tools, and manage vast amounts of project data, they will act as true copilots across all SDLC roles. The agentic mesh — a network of specialized AI agents—will coordinate and adapt autonomously to project needs, connecting quality assurance analysts, product managers, designers, architects, DevOps and database admins in a cohesive system that anticipates and meets workflow demands.

   With the agentic mesh, enterprises can deploy AI copilots that offer services such as high-fidelity UI testing, dynamic UX design, advanced prompting and domain-specific customization. This interconnected intelligence will automate routine tasks, allowing engineers to focus on high-value, innovation-driven initiatives. Ultimately, this will enable enterprises to deliver enhanced customer experiences, agile development processes and a new standard of efficiency and productivity.”

   Reply
10. February 20, 2025 at 11:22 am

    Tomi Engdahl says:

    Improve Business Productivity with Agentic Process Automation
    https://derobia.com/?utm_source=google&utm_medium=search&utm_campaign=search_022025&gad_source=1&gclid=EAIaIQobChMIiOOZuYvSiwMV7RaiAx22BCGbEAMYASAAEgIbNfD_BwE

    Get the job done by utilizing AI Agents that connect customers directly to your processes. Leverage your own data and the power of AI to improve work productivity. Benefit from different types of AI and a quick ROI with Derobia.

    How it works

    1. The customer sends a request in their preferred channel, which the agent captures.
    2. The agent analyzes and recognizes the content of the request with help from generative AI.
    3. By utilizing data from past incidents, the agent knows what it needs to do and gets the job done.

    Reply
11. February 20, 2025 at 11:24 am

    Tomi Engdahl says:

    Why Agentic AI is the Next Big Thing in the Business Industry
    https://www.testingxperts.com/blog/agentic-ai-in-business-industry/

    The evolution of the artificial intelligence (AI) domain is progressing at a rapid pace, and just when we are wrapping our heads around GenAI, another game-changing technology has come into the picture: agentic AI. It’s not just another buzzword; it has the capabilities to transform business processes by automating workflows and decision-making and even predicting customers’ needs. This technology brings the versatility and flexibility of LLMs and the accuracy of traditional programming together.

    Agentic AI is one of the innovative advancements in the AI industry. It is the combination of different AI approaches, techniques, and models that create a new series of autonomous agents to analyze data, establish goals, and create action plans to achieve them. And the plus point is that businesses do not even require a lot of human input to manage it. Compared to traditional AI models that simply execute predefined prompts or tasks, agentic AI can make decisions and plan actions on its own and can even learn from its experiences to fulfill the goals set by its users.

    Agentic AI employs advanced AI techniques, such as reinforcement learning, ML algorithms, and LLMs, to constantly learn and improve with every interaction. For example, LLMs leveraged by OpenAI’s ChatGPT, Meta Llama, and Google Gemini assist in making autonomous systems analyze, understand, and respond to NL (natural language) commands. These systems can also analyze data and identify patterns between different datasets. By doing so, they can learn from their actions and improve their decision-making capabilities.

    Agentic AI is navigating a new era by supporting autonomous agents with independence in non-rules-based processes and decisions, which will transform industries and the way we interact with technology. But it’s not just about automation; it’s about upskilling machines so that they can become partners of humans in solving complex problems, support businesses with critical thinking and decision-making capabilities, and take action and learn from them.

    Why Are Tech Leaders Focusing on Agentic AI?

    Following are some of the reasons why tech leaders are focusing on agentic AI solutions:

    In fast-paced and ever-evolving processes like supply chain management, cybersecurity, customer support, and finance/banking, agentic AI would allow businesses to adjust their strategies in real time and facilitate quick decision-making. This would allow businesses to remain adaptable and resilient to changing market conditions.

    Agentic AI solutions can solve complex problems by leveraging ML with goal-oriented behaviour to analyze data, identify data, and autonomously make decisions with optimal outcomes. It will also offer real-time problem-solving solutions in the dynamic business ecosystem.

    By leveraging agentic AI, businesses can streamline their decision-making processes and respond to changing market conditions in real time. The autonomous nature of agentic AI will allow this technology to quickly process large amounts of data, significantly speeding up workflows. This time efficiency helps businesses stay competitive by enabling them to act faster.

    Agentic AI systems can adapt to rapidly changing market conditions, making them highly scalable. As business demands grow or evolve, these systems will seamlessly adjust and continue functioning without requiring substantial new resource investments. This level of scalability will ensure businesses can expand operations without encountering the usual bottlenecks of scaling traditional processes.

    Some of the key characteristics of agentic AI are autonomy and handling tasks with approximately zero supervision. This level of independence would allow systems to handle complex workflows and challenges in real time, allowing human employees to initiate important business-development strategies and reduce operational bottlenecks. This is another benefit that tech leaders are looking for in terms of efficiency.

    Agentic AI can significantly lower operational costs by automating complex workflows and eliminating manual tasks. With reduced dependency on human supervision, businesses can save costs while improving accuracy and minimizing costly human errors. This cost-efficiency makes it an attractive solution for organizations aiming to optimize their resources.

    Agentic AI offers robust risk management capabilities by autonomously analyzing data and making decisions based on predefined goals. This technology has the ability to detect potential risks in real time, allowing businesses to mitigate them proactively. Understanding emerging threats or inefficiencies would allow tech leaders to make informed decisions that minimize risk and safeguard business operations.

    Enhancing Cybersecurity Measures:

    As digital threats have become a big problem for businesses, agentic AI can guard network security tirelessly. AI agents can autonomously monitor network traffic, identify loopholes, detect anomalies, and run remediation measures to attack against cyber threats in real time without human supervision. It can help organizations enhance their security infrastructure and address complex security challenges.
    Transforming Customer Support Service:

    Every business across all industries wants to deliver an expectational customer experience. Agentic AI can enhance customer support service by assisting software agents in providing personalized and 24/7 service beyond simple FAQs and automated responses. AI-enabled customer support agents would be able to understand written and oral queries, predict customer requirements, and resolve complex issues on their own. This level of hyper-personalization can help build brand loyalty and upscale customer experience.
    Integrating Agentic AI and IoT:

    It’s been noticed that various use cases are possible with the integration of Agentic AI and the Internet of Things (IoT). For instance, a network of interconnected sensors and devices equipped with an agentic-AI-powered agent can monitor, analyze, and optimize operations in real time. This can completely transform and take industries like healthcare, transformation, and manufacturing to a whole new level of improved efficiency, safety, and reduced costs.

    Ethical AI Considerations that Require Attention

    Although Agentic AI may have so many perks across industries, it also brings some of its challenges. Considering ethical dilemmas like ensuring the decision-making aligns with human values, the challenges are bound to occur. Because of the complex nature of AI models, businesses would face obstacles in understanding or interpreting their decision-making. Another ethical issue is ensuring accountability and trust in high-stake applications. Who will be held accountable if the agentic AI makes a blunder?

    Another ethical issue is data privacy and security. These systems will turn business processes autonomous and independent, so enterprises will need robust security measures to ensure protection against breaches and human misuse.

    Reply
12. February 20, 2025 at 11:25 am

    Tomi Engdahl says:

    The rise of Agentic AI: infrastructure as key to autonomy
    https://www.techzine.eu/news/infrastructure/128453/the-rise-of-agentic-ai-infrastructure-as-key-to-autonomy/

    Vultr announced that it is expanding its cloud offerings to support Internet of Things (IoT) solutions.

    Companies can easily integrate IoT systems into their cloud environment by combining a hybrid cloud infrastructure and a modular architecture. This ensures a secure, scalable, real-time data flow, crucial for developing Agentic AI workflows.

    Research by IDC FutureScape shows that every dollar invested in Agentic AI can yield a return of 3.7 times. Gartner also predicts that by 2028, about one-third of all enterprise software applications will incorporate Agentic AI, up sharply from less than one per cent in 2024. As a result, 15 per cent of daily business decisions will be made autonomously.

    Reply
13. February 20, 2025 at 11:26 am

    Tomi Engdahl says:

    Agentic AI
    https://www.uipath.com/ai/agentic-ai

    Learn about the future of work, where state-of-the-art AI and automation combine to create powerful autonomous agents able to understand, build, and perform complex business processes.

    What is agentic AI?

    Agentic AI is an emerging technology that is set to transform industries everywhere. It combines new forms of artificial intelligence (AI) like large language models (LLMs), traditional AI such as machine learning, and enterprise automation to create autonomous AI agents that can analyze data, set goals, and take actions with decreasing human supervision. These agents are capable of decision making and dynamic problem-solving, learning, and improving through every interaction.

    Agentic AI is a probabilistic technology with high adaptability to changing environments and events. It relies on patterns and likelihoods to make decisions and take actions, as opposed to deterministic systems—such as Robotic Process Automation (RPA)—that follow fixed rules and predefined outcomes. Agentic AI now makes it possible to automate many workflows and business processes that deterministic systems have not been capable of addressing on their own.

    Agentic AI doesn’t just enable enterprises to automate specific tasks—it powers intelligent systems capable of understanding context, adapting to new information, and collaborating with humans to solve complex challenges. By enabling machines to act independently in unstructured environments, agentic AI is redefining what automation can achieve.

    While agentic AI is opening up new areas to automation, RPA remains critical for highly compliant, secure and resilient operations. Therefore, the future of enterprise workflows will be a combination of both probabilistic and deterministic technologies, working together.

    Agentic AI and agentic automation

    Agentic AI has enabled a new type of automation—agentic automation—which can optimize complex, unstructured processes that traditional rules-based automation can’t address by itself. Agentic automation marks a shift from traditional, rules-based automation to a more dynamic, context-aware approach. While RPA has been instrumental in automating structured, repetitive tasks, agentic automation introduces the capability to handle complex, decision making processes that can adapt in real time. This isn’t just an expansion of RPA’s technological footprint; it’s a transformative approach that greatly complements existing RPA offerings, enhancing an enterprise’s ability to tackle larger, more complex automations.

    Agentic automation involves a symbiotic combination of AI agents, RPA robots, and people. People provide the goals for the agents, ensure governance, and step in when human judgment and review is required (human in the loop). Robots maximize the accuracy, productivity, and success of AI agents by collecting the data required for agents to make decisions (for example, logging in, connecting, and understanding information across multiple systems); they also can complete a wide range of other defined actions for agents.

    It is becoming increasingly clear that an orchestrated ecosystem of agents, robots, and people, managed on the same platform, offers higher productivity and better security and control.

    Reply
14. February 20, 2025 at 11:27 am

    Tomi Engdahl says:

    Agentic AI: The Next Big Thing in Telecommunications?
    https://www.juniperresearch.com/resources/blog/agentic-ai-the-next-big-thing-in-telecommunications/

    Agentic AI is an emerging concept in the field of artificial intelligence that has recently gained significant prominence. The technology promises to offer new capabilities, such as the ability to adapt, learn, and make independent decisions with minimal human oversight required.

    Agentic AI systems typically leverage multiple LLMs (large language models) that are deployed across various applications to enable autonomous decision making and action-taking and the completion of tasks with complex workflows.

    Whilst generative AI has been advertised as a tool to automate various functions within the telecommunications industry, this ability is limited by the data with which the AI system is provided. However, Agentic AI will fulfil an unmet need in the telecommunications sector, by automating processes that enhance operational efficiency across the industry.

    With the advancements that agentic AI can provide over both traditional AI and generative AI, Juniper Research anticipates that we will see it gain increased attention from players in the telecommunications industry over the next year. Autonomous decision making will enable better network management, customer service over chatbots, and more responsive security over networks.

    The autonomy of Agentic AI will ensure that it offers a clear benefit over other AI technologies. But what is the specific relevance of the technology for the telecommunications industry?

    Benefits of Agentic AI in Telecommunications

    There are several different ways in which agentic AI systems could be utilised in telecommunications; offering advanced solutions for both network operators and enterprises.

    For network operators, agentic AI will be used for the following:

    Network Management and Optimisation: Agentic AI could be used to optimise the performance of networks and reduce latency, by monitoring network traffic and automatically adjusting parameters. The predictive capabilities of agentic AI will also allow it to predict when a network failure will occur; in turn minimising downtime by ensuring proactive maintenance.

    Agentic AI will aid in the monetisation of 5G networks for real-time IoT applications. Agentic AI will not only help with the management and optimising of the growing number of applications connected to 5G networks, but could help to ensure that applications receive a consistent performance from the network.

    Network Security and Fraud Detection: Agentic AI’s ability to adapt based on its learning will allow it to provide enhanced network security solutions within the telecommunications industry. Where current AI systems can monitor patterns in network traffic and identify when a change in these patterns occurs, agentic AI can go one step further and adjust security protocols based on the threat that it identifies.

    For enterprises, agentic AI could be used for the following:

    Virtual Assistants: Agentic AI will enable virtual assistants to autonomously perform multi-step tasks; taking actions across multiple applications or services based on a user’s request. These virtual assistants can be used for both customer care and conversational commerce use cases, where they can make autonomous decisions based on customer requests and take action accordingly.

    For conversational commerce interactions, these virtual assistants will be able to handle shopping tasks, including the process of searching for products and completing the purchase on behalf of the customer, with appropriate authorisation; creating a more secure customer experience.

    Personalisation of Customer Experiences: Agentic AI also has the potential to predict when a customer will most likely respond to a messaging campaign. The technology will be able to predict changing user preferences, such as a preferred channel for communication, based on patterns and can adapt its approach without the need for human intervention.

    By employing agentic AI within business messaging applications, businesses will see increased engagement with messaging campaigns, as the technology will more accurately predict the channel with the highest engagement and the time to send a message to each customer.

    Telcos are already exploring and implementing AI technologies, however, there is not yet widespread adoption. Juniper Research anticipates that the first agentic AI solutions for telcos will become commercially available in 2026.

    Reply
15. February 20, 2025 at 11:29 am

    Tomi Engdahl says:

    Executive Platform
    Agentic AI – A Game Changer for IT
    https://blogs.cisco.com/news/agentic-ai-a-game-changer-for-it

    AI-driven innovations are beginning to change every aspect of our lives from home to school to work. Generative AI can augment our work such as an AI Assistant that responds to prompts, or it can enhance or automate activities in an almost imperceptible way. When we look back 5 or so years from now, we will be blown away by how much has changed – highways and streets full of cars that safely drive themselves, robots that make dinner and do the dishes, and networks that fully heal themselves end-to-end. Generative AI is getting smarter and more powerful.
    The Future of Generative AI is Agentic

    With the advent of agentic AI, assistance is moving to automation where systems act with agency to achieve specific goals. Agentic AI systems can make rapid decisions, manage complex tasks, and adapt to changing conditions. They have agency to reach beyond the data their large language model was trained on and interact with external environments such as IOT sensors, cloud platforms or analytics software. The possibilities are endless for what an agentic AI system can achieve in improving customer experiences, increasing productivity, and creating new innovation.

    In a recent report on the top 10 strategic technology trends for 2025, Gartner predicts that by 2028, at least 15% of day-to-day work decisions will be taken autonomously through agentic AI. And because of its ability to take action and make decisions, agentic AI could help entire industries reimagine the way work gets done. For example, agentic systems can write software code to solve a problem, create a human-like gaming character, or act as a medical research assistant​. The opportunity to innovate has never been greater.

    Helping IT Deliver Better Experiences

    Leveraging various AI models and approaches including agentic AI in the network, new capabilities can be enabled to do things like analyze data, detect network anomalies, allocate network resources, and optimize network quality —all with little to no human supervision. This can be a game-changer for IT. The technology can help automate routine tasks, identify and solve issues, and optimize performance, resulting in transformative business outcomes like reduced downtime and improved user experiences. This technology can also lend to a more robust network infrastructure that adapts to changing demands and scales as your business grows.
    Delivering ​Infrastructure for AI​ with Resilience

    According to the Cisco AI Readiness Index, 54% of those surveyed acknowledge their infrastructure has limited or moderate scalability and flexibility to accommodate the increasing needs for AI-powered technologies. Adaptable, secure, and resilient infrastructure gets your data center ready for the power of agentic AI. 

    Reply
16. February 20, 2025 at 12:11 pm

    Tomi Engdahl says:

    Artificial intelligence (AI), once limited to simple question-answer formats, has now advanced to agents capable of performing tasks as efficiently as humans. These agents have also far surpassed virtual assistants like Siri and Alexa, showing significant potential in drug discovery in healthcare, fraud detection in finance, supply chain optimization in e-commerce, and so much more.

    A Capgemini survey of 1,100 executives from major enterprises found that 10% of firms already use AI agents, and 82% plan to incorporate them within the next three years. Notably, 60% intend to build AI agents within a year, while a quarter foresee longer timelines.

    To stay competitive, engage customers, and increase profits, it’s time for businesses to integrate this transformative technology. The question now is: How to build an AI agent?

    AI agent building
    https://www.servicenow.com/products/ai-agents.html

    How to Build AI Agents for Beginners (2025)
    https://botpress.com/blog/build-ai-agent

    9 Best AI Chatbot Platforms: A Comprehensive Guide (2025)
    https://botpress.com/blog/9-best-ai-chatbot-platforms

    A Step-by-Step Guide to Building an AI Agent From Scratch
    https://www.neurond.com/blog/how-to-build-an-ai-agent

    Using pre-built frameworks streamlines the development process by offering ready-made components for common AI agent functions. These frameworks often integrate advanced language models to handle core tasks. Some popular options are:

    Microsoft Autogen: Known for its collaborative features and simplified agent-building process.

    LangChain: An open-source framework offering a modular architecture for building agents.

    LlamaIndex: Best for complex tasks related to information retrieval.

    crewAI: A paid platform with pre-built components and tools for creating AI assistants.

    How to Build an AI Agent
    https://www.salesforce.com/agentforce/build-ai-agent/

    Build and Recruit AI Agents
    Build and recruit teams of AI agents to complete tasks on autopilot.
    https://relevanceai.com/agents

    Reply
17. February 20, 2025 at 12:13 pm

    Tomi Engdahl says:

    I found where to build AI agents (free and paid) Easy
    https://www.reddit.com/r/SaaS/comments/1cndcuy/i_found_where_to_build_ai_agents_free_and_paid/

    Where to Build My Own AI Agents: The Top Free and Top Paid Options

    To build your own AI Agents with no coding, you need to use a platform. Luckily, there are multiple free and paid options.

    These platforms cater to different levels of expertise, from beginners to advanced developers, and offer a range of functionalities to suit various project requirements.

    Outside of google’s vertex AI, there are other options. Here’s a breakdown of the top free and paid options available in 2024.

    Build AI Agents for Free

    To build your own AI Agents for free to do whatever you want, you have a couple of options on which platform you use.

    They all work, but there are pros and cons to each one. Here is a brief overview of each one to help you decide

    1.

    Zapier Central to creat your own AI Agents

    Zapier Central: A new no-code AI Agent builder that integrates with Zapier’s extensive catalogue of apps, allowing for easy creation of AI agents without coding knowledge. It’s particularly noted for its user-friendly interface and the ability to connect live data for dynamic interactions.

    2. Agentgpt – New tool to create your own AI Agents

    Agentgpt: Supported by an open-source community, AgentGPT offers a web-based platform that enables users to build and deploy AI agents directly from their browsers. It’s designed to be accessible to a wide audience, allowing for the creation of autonomous agents with the ability to learn and take action.

    3. AutoGPT – New tool to create your own specialized ai agents

    AutoGPT: An experimental, open-source autonomous AI agent built upon the GPT-4 language model. AutoGPT autonomously links together tasks to accomplish a user-defined overarching goal, offering a high degree of customization and flexibility for developers.

    4. GPTPilot (now Pythagora) A tool to create your own ai agents

    GPTPilot (now Pythagora): An open-source AI agent that can build full-stack applications. It’s capable of adding features to existing projects and is completely free. The platform supports various open-source models, including GPT-4, and can be run locally.

    5. Autonomous Virtual Agents (AVAs): Offered by my-ava.net.

    A tool to create your own ai agents Autonomous Virtual Agents (AVAs): Offered by my-ava.net, these agents feature the latest functionalities of ChatGPT plus multiplatform integration. They come with dynamic avatars, voice chat, and a “sentience core” for enhanced interaction capabilities. The software is open-source and encourages community contributions.

    Build AI Agents with Paid Options

    If you’re looking for the best possible tools out there to build your AI Agents, you may have to pay a bit.

    Here are the top platforms to build your AI Agents that cost a bit.

    Reply
18. February 20, 2025 at 12:16 pm

    Tomi Engdahl says:

    AgentGPT
    https://agentgpt.reworkd.ai/

    Free Trial

    A small taste of what AgentGPT offers.

    $0
    / month

    5 demo agents a day using GPT-3.5-Turbo
    Limited plugin integrations
    Limited web search capabilities

    Reply
19. February 20, 2025 at 12:17 pm

    Tomi Engdahl says:

    Microchip toi tekoälyassistentin mikro-ohjainkehitykseen
    https://etn.fi/index.php/13-news/17177-microchip-toi-tekoaelyassistentin-mikro-ohjainkehitykseen

    Microchip Technology on ottanut tekoälyn hyötykäyttöön auttaakseen ohjelmistokehittäjiä ja sulautettujen järjestelmien insinöörejä koodin kirjoittamisessa ja virheiden korjaamisessa. Yhtiö on julkaissut uuden MPLAB AI Coding Assistant -työkalun, joka on ilmainen Microsoft Visual Studio Code -laajennus.

    Työkalu perustuu suosittuun avoimen lähdekoodin Continue-assistenttiin ja sisältää Microchipin oman AI-pohjaisen chatbotin reaaliaikaista tukea varten.

    Reply
20. February 20, 2025 at 1:41 pm

    Tomi Engdahl says:

    Free local AI Agents with Ollama
    https://www.youtube.com/watch?v=rZuj-5e02J8

    Introducing the NetBox AI Agent—a fully local, private AI assistant powered by:
    Ollama for seamless AI model hosting
    Llama 3.1 (Meta) for natural language processing
    Command-r7b (Cohere) for powerful task execution

    No RAG, No Vector Stores, No Embeddings.
    This is pure, lightweight, and powerful AI—running locally on your own hardware.

    https://github.com/automateyournetwork/netbox_react_agent

    Reply
21. February 20, 2025 at 1:42 pm

    Tomi Engdahl says:

    What is MCP? Integrate AI Agents with Databases & APIs
    https://www.youtube.com/watch?v=eur8dUO9mvE

    Unlock the secrets of MCP!
    Dive into the world of Model Context Protocol and learn how to seamlessly connect AI agents to databases, APIs, and more. Roy Derks breaks down its components, from hosts to servers, and showcases real-world applications. Gain the knowledge to revolutionize your AI projects!

    Reply
22. February 20, 2025 at 1:44 pm

    Tomi Engdahl says:

    GraphRAG: The Marriage of Knowledge Graphs and RAG: Emil Eifrem
    https://www.youtube.com/watch?v=knDDGYHnnSI

    A famous poet once said “Natural language is most powerful when it can draw from a rich context.” Ok fine, I said that. But that’s true of both poetry, and of LLMs! Well, Knowledge Graphs excel at capturing context. How can combining Knowledge Graphs with RAG – an emerging technique known as GraphRAG – give context to your RAG application, and lead to more accurate and complete results, accelerated development, and explainable AI decisions? This talk will go deep on the why and how of GraphRAG, and where best to apply it. You’ll get concepts, examples, and specifics on how you can get started. You’ll walk away with an understanding of how GraphRAG can improve the context you pass to the LLM and the performance of your AI applications.

    Reply
23. February 20, 2025 at 5:41 pm

    Tomi Engdahl says:

    “OpenAI’s o1-preview tried to cheat 37% of the time; while DeepSeek R1 tried to cheat 11% of the time”

    When AI Thinks It Will Lose, It Sometimes Cheats, Study Finds
    https://time.com/7259395/ai-chess-cheating-palisade-research/?fbclid=IwY2xjawIkQ5FleHRuA2FlbQIxMQABHXnIFSZw_ilPVJZUhjnqV8dQRBBPyzhBgFApQxglT4vEzjSxvKog7yoA_g_aem_AHke9L5KWkjWNLWid80QoA

    The models’ enhanced ability to discover and exploit cybersecurity loopholes may be a direct result of powerful new innovations in AI training, according to the researchers. The o1-preview and R1 AI systems are among the first language models to use large-scale reinforcement learning, a technique that teaches AI not merely to mimic human language by predicting the next word, but to reason through problems using trial and error. It’s an approach that has seen AI progress rapidly in recent months, shattering previous benchmarks in mathematics and computer coding. But the study reveals a concerning trend: as these AI systems learn to problem-solve, they sometimes discover questionable shortcuts and unintended workarounds that their creators never anticipated, says Jeffrey Ladish, executive director at Palisade Research and one of the authors of the study. “As you train models and reinforce them for solving difficult challenges, you train them to be relentless,” he adds.

    Reply
24. February 20, 2025 at 8:06 pm

    Tomi Engdahl says:

    Continue
    https://www.continue.dev/

    The leading open-source AI code assistant. You can connect any models and any context to create custom autocomplete and chat experiences inside the IDE

    Reply
25. February 20, 2025 at 8:36 pm

    Tomi Engdahl says:

    Llama Raspberry Pi

    Run Llama on your Raspberry Pi 5 without using Ollama
    https://medium.com/@wesselbraakman/run-llama-on-your-raspberry-pi-5-without-using-ollama-7ebc128ff34e

    So I have been tinkering with my Raspberry Pi 5 8gb since I got it in december. I found many guides to install an LLM on it, but kept running into issues that I could not easily get past. A lot of this had to do with the source computer on which I was supposed to retrieve/build/quantize the LLM, and some of this had to do with me not being able to install all I needed on my RPi5 without running into issues.

    It is therefore I am writing this guide, in which I pinpoint where I got stuck, and write how I worked around this.

    So this is by no means a guide that I have completely figured out on my own (I am not an expert on any of these topics), but more a guide that should help out in case someone gets stuck.

    What do we need?

    Source PC with either Windows or a Linux distribution to retrieve and quantize the LLM(‘s)
    8GB Raspberry Pi 5 to run the LLM on
    A memory card with at least 32GB containing a pre-installed OS such as Raspbian (I personally use Ubuntu 23.04 on my RPi for this)
    An USB stick with at least 22GB of space available to transport the LLM from your source PC to your RPi

    Downloading and building the Llama project

    For downloading the Llama project into our workspace, we will use the “git clone” command.

    > git clone https://github.com/ggerganov/llama.cpp

    After downloading is finished, we will enter the folder we just downloaded

    Reply
26. February 20, 2025 at 8:38 pm

    Tomi Engdahl says:

    How to Run a ChatGPT-like AI Bot on a Raspberry Pi
    https://github.com/garyexplains/examples/blob/master/how-to-run-llama-cpp-on-raspberry-pi.md

    Here is my step-by-step guide to running Large Language Models (LLMs) using llama.cpp on a Raspberry Pi. These instructions accompany my video How to Run a ChatGPT-like AI on Your Raspberry Pi.

    Performance

    The lower the quantization, the better the performance, but the lower the accuracy. An interesting test is whether a higher quantized 7B model is more accurate than a lower quantized 13B model. Let me know if you find out!

    For a Raspberry Pi you should stick with Q4 or lower.

    Reply
27. February 20, 2025 at 8:39 pm

    Tomi Engdahl says:

    How to run the new Llama 3.1 on Raspberry Pi!!!
    https://www.youtube.com/watch?v=KcWKTdkUpoQ

    In this Tutorial, You’ll learn how to run Llama 3.1 on Raspberry Pi 5.

    We are going to use a method called Llamafile to do run Llama 3.1 on RPi5.

    Llamafile is an executable file for distributing LLMs.

    As part of this, we’ll run Llama 3.1
    1. Using GUI
    2. Using HTTP Endpoint as a CuRL command!

    It’s pretty insane how far we have come from needing large GPUs to run LLMs to Raspberry Pi!

    Timestamp

    00:00 Intro
    00:36 Inside Raspberry Pi
    02:38 Running LLama 3.1 inside Raspberry Pi
    04:40 Calling Llama 3.1 via CuRL (HTTP Endpoint)
    06:11 Using Llama 3.1 with Llama CPP GUI

    https://github.com/Mozilla-Ocho/llamafile

    Reply
28. February 20, 2025 at 8:40 pm

    Tomi Engdahl says:

    How to install and Run Llama 3.2 1B and 3B LLMs on Raspberry Pi and Linux Ubuntu
    https://www.youtube.com/watch?v=ql7aXdQ3-68

    In this Large Language Model (LLM) and machine learning tutorial, we explain how to run Llama 3.2 1B and 3B LLMs on Raspberry Pi in Linux Ubuntu. In this tutorial, we use Raspberry Pi 4. However, the performance and speed of running the models will be better on Raspberry Pi 5. Almost everything explained in this tutorial applies to Raspberry Pi 5. The only difference is that the process of overclocking the processor will be different since you need to use different settings and parameters that are more suitable for Raspberry Pi 5.

    Before we start with explanations, we need to emphasize the following:
    (try to carefully listen to what is explained here)

    In this tutorial, we will be using Raspberry Pi 4 with 4GB of RAM. To enhance the performance of Raspberry Pi 4, we will overclock its GPU and CPU. Furthermore, we will increase the swap memory file size in order to be able to run 3B model. This is very important otherwise, we will not be able to run 3B model since it cannot fit in our memory. On the other hand, if you are using Raspberry Pi 4 with 8GB RAM, this might not be necessary. However, we suggest everyone to increase the swap memory size. This will increase the stability and make sure that the applications do not stop due to the lack of RAM memory. On the other hand, if you are using Raspberry Pi 5, you can also try to increase the swap memory. Here is the disclaimer regarding overclocking and swap memory adjustment:

    Reply
29. February 20, 2025 at 8:41 pm

    Tomi Engdahl says:

    Running DeepSeek R1 Locally on a Raspberry Pi
    https://dev.to/jeremycmorgan/running-deepseek-r1-locally-on-a-raspberry-pi-1gh8

    DeepSeek R1 shook the Generative AI world, and everyone even remotely interested in AI rushed to try it out. It is a great model, IMO. As you may know, I love to run models locally, and since this is an open-source model, of course, I had to try it out. It works great on my Mac Studio and 4090 machines.

    Running Deepseek on a Raspberry Pi
    https://pimylifeup.com/raspberry-pi-deepseek/

    Reply
30. February 20, 2025 at 8:43 pm

    Tomi Engdahl says:

    GitHub Copilot in VS Code
    https://code.visualstudio.com/docs/copilot/overview

    Announcing a free GitHub Copilot for VS Code
    December 18, 2024 by Burke Holland, @burkeholland
    https://code.visualstudio.com/blogs/2024/12/18/free-github-copilot

    We’re excited to announce an all new free plan for GitHub Copilot, available for everyone today in VS Code. All you need is a GitHub account. No trial. No subscription. No credit card required.

    With GitHub Copilot Free you get 2000 code completions/month. That’s about 80 per working day – which is a lot. You also get 50 chat requests/month, as well as access to both GPT-4o and Claude 3.5 Sonnet models.

    If you hit these limits, ideally it’s because Copilot is doing its job well, which is to help you do yours! If you find you need more Copilot, the paid Pro plan is unlimited and provides access to additional models like o1 and Gemini (coming in the new year).

    https://github.com/features/copilot/plans?cft=copilot_lo.features_copilot

    Reply
31. February 20, 2025 at 8:44 pm

    Tomi Engdahl says:

    https://www.continue.dev/

    Reply
32. February 20, 2025 at 8:44 pm

    Tomi Engdahl says:

    https://www.reddit.com/r/vscode/comments/18qvxqc/which_free_ai_coding_assistant_do_you_suggest_on/

    Reply
33. February 21, 2025 at 10:53 am

    Tomi Engdahl says:

    AI Can Supercharge Productivity, But We Still Need a Human-in-the-Loop
    https://www.securityweek.com/ai-can-supercharge-productivity-but-we-still-need-a-human-in-the-loop/

    AI systems can sometimes struggle with complex or nuanced situations, so human intervention can help identify and address potential issues that algorithms might not.

    Reply
34. February 21, 2025 at 10:55 am

    Tomi Engdahl says:

    Benj Edwards / Ars Technica:
    Microsoft researchers introduce Magma, an AI foundation model that combines visual and language processing to control software interfaces and robotic systems — On Wednesday, Microsoft Research introduced Magma, an integrated AI foundation model that combines visual and language processing …

    Microsoft’s new AI agent can control software and robots
    Magma could enable AI agents to take multistep actions in the real and digital worlds.
    https://arstechnica.com/ai/2025/02/microsofts-new-ai-agent-can-control-software-and-robots/

    On Wednesday, Microsoft Research introduced Magma, an integrated AI foundation model that combines visual and language processing to control software interfaces and robotic systems. If the results hold up outside of Microsoft’s internal testing, it could mark a meaningful step forward for an all-purpose multimodal AI that can operate interactively in both real and digital spaces.

    Microsoft claims that Magma is the first AI model that not only processes multimodal data (like text, images, and video) but can also natively act upon it—whether that’s navigating a user interface or manipulating physical objects. The project is a collaboration between researchers at Microsoft, KAIST, the University of Maryland, the University of Wisconsin-Madison, and the University of Washington.

    Magma: A Foundation Model for Multimodal AI Agents
    https://microsoft.github.io/Magma/

    Magma is the first foundation model that is capable of interpreting and grounding multimodal inputs within its environment. Given a described goal, Magma is able to formulate plans and execute actions to achieve it. By effectively transferring knowledge from freely available visual and language data, Magma bridges verbal, spatial and temporal intelligence to navigate complex tasks and settings.

    Reply
35. February 21, 2025 at 10:57 am

    Tomi Engdahl says:

    We’ve seen other large language model-based robotics projects like Google’s PALM-E and RT-2 or Microsoft’s ChatGPT for Robotics that utilize LLMs for an interface. However, unlike many prior multimodal AI systems that require separate models for perception and control, Magma integrates these abilities into a single foundation model.

    Google’s PaLM-E is a generalist robot brain that takes commands
    ChatGPT-style AI model adds vision to guide a robot without special training.
    https://arstechnica.com/information-technology/2023/03/embodied-ai-googles-palm-e-allows-robot-control-with-natural-commands/

    Google’s RT-2 AI model brings us one step closer to WALL-E
    “First-of-its-kind” robot AI model can recognize trash and perform complex actions.
    https://arstechnica.com/information-technology/2023/07/googles-rt-2-ai-model-brings-us-one-step-closer-to-wall-e/

    Robots let ChatGPT touch the real world thanks to Microsoft
    A new API allows ChatGPT to control robots through natural language commands.
    https://arstechnica.com/information-technology/2023/02/robots-let-chatgpt-touch-the-real-world-thanks-to-microsoft/

    Last week, Microsoft researchers announced an experimental framework to control robots and drones using the language abilities of ChatGPT, a popular AI language model created by OpenAI. Using natural language commands, ChatGPT can write special code that controls robot movements. A human then views the results and adjusts as necessary until the task gets completed successfully.

    The research arrived in a paper titled “ChatGPT for Robotics: Design Principles and Model Abilities,” authored by Sai Vemprala, Rogerio Bonatti, Arthur Bucker, and Ashish Kapoor of the Microsoft Autonomous Systems and Robotics Group.

    https://www.microsoft.com/en-us/research/uploads/prod/2023/02/ChatGPT___Robotics.pdf

    ChatGPT for Robotics
    https://www.youtube.com/watch?v=NYd0QcZcS6Q

    Reply
36. February 21, 2025 at 10:58 am

    Tomi Engdahl says:

    Steve Lohr / New York Times:
    Software engineers, academics, and others say AI coding tools will likely prompt an evolution rather than extinction, pushing developers to learn new skills

    A.I. Is Prompting an Evolution, Not Extinction, for Coders
    https://www.nytimes.com/2025/02/20/business/ai-coding-software-engineers.html?unlocked_article_code=1.yk4.XrJ-.4YoaKmXX9SEh&smid=url-share

    A.I. tools from Microsoft and other companies are helping write code, placing software engineers at the forefront of the technology’s potential to disrupt the work force.

    John Giorgi uses artificial intelligence to make artificial intelligence.

    The 29-year-old computer scientist creates software for a health care start-up that records and summarizes patient visits for doctors, freeing them from hours spent typing up clinical notes.

    To do so, Mr. Giorgi has his own timesaving helper: an A.I. coding assistant. He taps a few keys and the software tool suggests the rest of the line of code. It can also recommend changes, fetch data, identify bugs and run basic tests. Even though the A.I. makes some mistakes, it saves him up to an hour many days.

    “I can’t imagine working without it now,” Mr. Giorgi said.

    That sentiment is increasingly common among software developers, who are at the forefront of adopting A.I. agents, assistant programs tailored to help employees do their jobs in fields including customer service and manufacturing. The rapid improvement of the technology has been accompanied by dire warnings that A.I. could soon automate away millions of jobs — and software developers have been singled out as prime targets.

    But the outlook for software developers is more likely evolution than extinction, according to experienced software engineers, industry analysts and academics. For decades, better tools have automated some coding tasks, but the demand for software and the people who make it has only increased.

    Reply
37. February 21, 2025 at 10:59 am

    Tomi Engdahl says:

    Bloomberg:
    Alibaba CEO Eddie Wu calls the pursuit of AGI the company’s “primary objective”, saying “our first and foremost goal is to pursue AGI” — “Our first and foremost goal is to pursue AGI,” Eddie Wu told investors on a call after the company reported results that surpassed analyst estimates.

    Alibaba CEO Wu Says AGI Is Now Company’s ‘Primary Objective’
    https://www.bloomberg.com/news/articles/2025-02-20/alibaba-ceo-wu-says-agi-is-now-company-s-primary-objective

    Reply
38. February 21, 2025 at 11:04 am

    Tomi Engdahl says:

    Gergely Orosz / The Pragmatic Engineer:
    A look at the potential reasons why software developer job listings on Indeed hit a five-year low in January, down 35% from 2020, including the impact of AI — Hi, this is Gergely with a bonus issue of the Pragmatic Engineer Newsletter. In every issue, I cover topics related to Big Tech …

    Software engineering job openings hit five-year low?
    https://blog.pragmaticengineer.com/software-engineer-jobs-five-year-low/

    Reply
39. February 21, 2025 at 11:05 am

    Tomi Engdahl says:

    Dean Takahashi / VentureBeat:
    Nvidia launches Signs, a new AI platform to teach American Sign Language and create a validated video library of ASL signs for learners and app developers

    Nvidia helps launch AI platform for teaching American Sign Language
    https://venturebeat.com/games/nvidia-helps-launch-ai-platform-for-teaching-american-sign-language/

    Nvidia has unveiled a new AI platform for teaching people how to use American Sign Language to help bridge communication gaps.

    The Signs platform is creating a validated dataset for sign language learners and developers of ASL-based AI applications.

    It so happens that American Sign Language is the third most prevalent language in the United States —
    but there are vastly fewer AI tools developed with ASL data than data representing the country’s most common languages, English and Spanish.

    Reply
40. February 21, 2025 at 11:07 am

    Tomi Engdahl says:

    Kate Rooney / CNBC:
    OpenAI COO Brad Lightcap says ChatGPT crossed 400M weekly active users in February 2025, up 33% from 300M in December 2024, and 2M paying enterprise customers

    OpenAI tops 400 million users despite DeepSeek’s emergence
    https://www.cnbc.com/2025/02/20/openai-tops-400-million-users-despite-deepseeks-emergence.html

    The artificial intelligence company saw 400 million weekly active users, up 33% in less than three months, OpenAI’s chief operating officer, Brad Lightcap, told CNBC.
    The growth came amid more competition from open source models like DeepSeek.
    “There’s an overall effect of people really wanting these tools, and seeing that these tools are really valuable,” Lightcap said.

    OpenAI appears to be growing quickly despite increasing competition.

    The San Francisco-based tech company had 400 million weekly active users as of February, up 33% from 300 million in December, the company’s chief operating officer, Brad Lightcap, told CNBC. These numbers have not been previously reported.

    Lightcap pointed to the “natural progression” of ChatGPT as it becomes more useful and familiar to a broader group of people.

    “People hear about it through word of mouth. They see the utility of it. They see their friends using it,” Lightcap said in an interview, adding that it takes time for individuals to find use cases that resonate. “There’s an overall effect of people really wanting these tools, and seeing that these tools are really valuable.”

    OpenAI is seeing that spill over to its growing enterprise business. The company now has 2 million paying enterprise users, roughly doubling from September, said Lightcap, pointing out that often employees will use ChatGPT personally and suggest to their companies that they implement the tool.

    “We get a lot of benefits, and a tail wind from the organic consumer adoption where people already have familiarity with the product,” he said. “There’s really healthy growth, on a different curve.”

    Developer traffic has also doubled in the past six months, quintupling for the company’s “reasoning” model o3, according to Lightcap. Developers use OpenAI to integrate the technology into their own applications. OpenAI counts Uber
    , Morgan Stanley, Moderna and T-Mobile among some of its largest enterprise customers.

    Reply
41. February 21, 2025 at 11:08 am

    Tomi Engdahl says:

    Jess Weatherbed / The Verge:
    Spotify begins accepting AI-narrated audiobooks recorded using ElevenLabs’ software; Spotify already allows AI-recorded audiobooks, with several restrictions

    Spotify is making it easier to release audiobooks narrated by AI
    https://www.theverge.com/news/616355/spotify-audiobooks-elevenlabs-ai-narration
    A new ElevenLabs partnership aims to help authors bring AI-narrated audiobooks to Spotify listeners

    Reply
42. February 21, 2025 at 1:48 pm

    Tomi Engdahl says:

    Suomalaismalli osaa arvioida GenAI-mallien tuotoksia tehokkaasti
    https://etn.fi/index.php/13-news/17181-suomalaismalli-osaa-arvioida-genai-mallien-tuotoksia-tehokkaasti

    Suomalainen tekoälyteknologia ottaa harppauksen eteenpäin, kun Root Signals -yritys julkaisee uuden Root Judge -mallin, joka on suunniteltu arvioimaan suurten kielimallien (LLM) luotettavuutta ja laatua. Tämä avoimen lähdekoodin malli tuo tarkkuutta ja läpinäkyvyyttä generatiivisten AI-järjestelmien arviointiin ja optimointiin.

    Root Judge on kehitetty Meta Llama-3.3-70B-Instruct -mallin pohjalta ja se pystyy tunnistamaan hallusinaatioita, vertailemaan erilaisten kielimallien tuotoksia sekä varmistamaan, että AI-ratkaisut täyttävät korkeat luotettavuusvaatimukset. Tämä tekee siitä arvokkaan työkalun yrityksille ja tutkijoille, jotka haluavat rakentaa turvallisempia ja tehokkaampia tekoälysovelluksia.

    Erityisesti Retrieval-Augmented Generation (RAG) -teknologian kanssa käytettäväksi suunniteltu Root Judge pystyy havaitsemaan virheellisiä tai harhaanjohtavia tietoja ja estämään niiden leviämisen. Lisäksi malli tukee parivertailuarviointeja, jotka auttavat käyttäjiä tekemään parempia päätöksiä tekoälyn tulosten perusteella.

    Kuten suomalaisen SiloAI:n avoimet Poro- ja Viking-mallit, Root Signalsin Root Judge on koulutettu LUMI-supertietokoneessa Kajaanissa.

    Reply
43. February 21, 2025 at 2:10 pm

    Tomi Engdahl says:

    Grok 3 – Onko tässä paras tekoäly?
    https://www.youtube.com/watch?v=I3dezqndWmQ

    Reply
44. February 21, 2025 at 2:35 pm

    Tomi Engdahl says:

    Troubleshooting Packets with AI
    https://www.youtube.com/watch?v=kYSUy10Y5U8

    Join us as we discuss AI-powered troubleshooting with Anand Ravi and Johnny Ghibril of B-Yond.

    Reply
45. February 21, 2025 at 2:38 pm

    Tomi Engdahl says:

    AI, Machine Learning, Deep Learning and Generative AI Explained
    https://www.youtube.com/watch?v=qYNweeDHiyU

    Join Jeff Crume as he dives into the distinctions between Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), and Foundation Models and how these technologies have evolved over time. He also explores the latest advancements in Generative AI, including large language models, chatbots, and deepfakes – and clarifies common misconceptions, simplifies complex concepts, and discusses the impact these technologies have on various fields.

    Reply
46. February 21, 2025 at 2:39 pm

    Tomi Engdahl says:

    DeepSeek, TikTok, Temu: How China is taking the lead in tech – BBC World Service
    https://www.youtube.com/watch?v=z7do1hhb6fE

    Reply
47. February 21, 2025 at 2:41 pm

    Tomi Engdahl says:

    AGILITY Demo: Transforming Network Troubleshooting & Analysis
    https://www.youtube.com/watch?v=qNlziEWx838

    Welcome to our AGILITY Demo video! B-Yond is proud to present AGILITY, a groundbreaking AI and ML-powered tool designed to revolutionize network troubleshooting and analysis in the telecommunications industry.

    In this demo, we’ll walk you through the key features of AGILITY, including its intuitive navigation, interactive sequence diagram, and protocol level analysis. You’ll see firsthand how AGILITY simplifies complex tasks, allowing network operators to quickly identify and resolve issues.

    AGILITY is more than just a tool; it’s a solution that enhances the efficiency of network troubleshooting and analysis, ensuring the delivery of reliable and high-quality services to customers.

    Ready to experience the future of network operations? Watch this demo to see AGILITY in action and discover how it can transform your network troubleshooting and analysis processes.

    Reply
48. February 22, 2025 at 6:11 am

    Tomi Engdahl says:

    Why AI Safety Researchers Are Worried About DeepSeek
    https://time.com/7210888/deepseeks-hidden-ai-safety-warning/

    Reply
49. February 22, 2025 at 8:11 am

    Tomi Engdahl says:

    Neuro-inspired AI framework uses reverse-order learning to enhance code generation
    https://techxplore.com/news/2025-02-neuro-ai-framework-reverse-code.html#google_vignette

    Reply
50. February 22, 2025 at 8:12 am

    Tomi Engdahl says:

    https://www.eetimes.com/startup-positron-takes-on-nvidia-with-fpgas/

    Reply