Yazar: ayaksız

The downside of AI has dominated the media, reaching extreme scenarios including the extinction of the human race. No one knows whether the threats presented by artificial intelligence are realistic, exaggerated, or something in between. However, in the meantime, the positive side of artificial intelligence is waiting to be discovered. Confidant: AI isn’t human, but it can serve as a sounding board for your deepest personal issues, including things you wouldn’t even disclose to a friend. A chatbot like ChatGPT will dialogue with you in a non-threatening way, patiently listening to everything you have to say, and offering specific answers…

Manufacturing operations generate vast amounts of data at every step in the process, from the design of products to the performance of assembly lines. It is, therefore, a vast challenge to make sense of this data – turning it into actionable insights that will benefit the business. Significant advances in two areas of technology – digital twins and AI – in manufacturing is releasing some of the constraints that firms faced when trying to use data effectively. “Manufacturers are no strangers to using data in everyday processes and procedures, but often the data is constrained to specific systems or processes…

Effective communication and collaboration are critical to driving strong performance and productivity in a hybrid working environment. Many people are unaware they can leverage data from within Microsoft Teams to support their analysis of employee productivity. To understand if a hybrid working model is effective, it is important to first identify what best practice looks like across roles and departments. Reviewing optimal working profiles enables organisations to identify key metrics for measurement. This is exactly what our customer Osotspa did using our application, Analytics 365 Collaboration Analytics. The Thailand-based beverage firm used Analytics 365 to review its hybrid working policy,…

Being a member of Generation X, I grew up watching linear TV and reading offline newspapers. My social interaction with friends was mainly face to face. However, I have rarely seen my children watch traditional TV shows or read an offline paper. Watching them grow up, I realise the importance social networks and video games have in their everyday lives, particularly in terms of staying in touch with their friends. This illustrates how, in the ever-evolving media and entertainment landscape, the transition from traditional TV consumption to the metaverse can captivate the hearts and minds of a new generation of…

The past decade saw organisations rapidly expanding their security teams and toolsets to handle the rise in threat event frequency. As these organisations enter 2024 however, they are finding that their security processes are now outdated, their toolset has become bloated, and their security teams are struggling to withstand cyberattacks. The time has come to hit reset and modernise security operations. Kyndryl’s Security Operations as a platform (SOaap) helps address these issues by creating a consolidated view of security tools, combined with automation, to optimise decision-making by security teams. SOaap is powered by Kyndryl Bridge, an open integration platform that…

Top reasons for failure The reasons that the failures occur vary a great deal. The top four that I see include: Inadequate architecture. Too often, businesses migrate to the cloud without adequate planning or understanding of cloud computing. Significant performance or reliability issues can arise from this. Or more likely, grossly underoptimized systems in the cloud that eat 5 to 10 times more money than they should. We’ve beaten those issues to death here, and I won’t dwell on it. Poorly defined service-level agreements (SLAs). Why do expected performance standards go unmet? It’s mainly due to ill-defined SLAs between the…

Azure AI Studio model catalog Azure AI Studio hosts AI models from Microsoft Research, OpenAI, Meta, Hugging Face, and Databricks, as well as NVIDIA base models, so that you can find the current best model for your application, or at least one that works well enough. In addition, Azure AI Studio offers half a dozen Azure OpenAI language models, some of which have fine-tuning capabilities. In general, the OpenAI models are offered “as a service,” meaning that they are deployed in a model pool with its own GPUs. When you provision them, you get an inference endpoint in your own…

Buy a used or new car recently? If you have, you understand the issues with “junk fees” that some dealers insist on tacking on. Junk fees refer to additional charges that car dealers may include in the purchase agreement. These fees are often unrelated to the actual cost of the car, and they can significantly inflate the overall price. I’m sure you’ve seen them before: administrative fees, document processing, and my favorite, advertising fees, meaning they want you to pay for their marketing expenses. They have some nerve to try that one, but it happens all the time. Hidden cloud…

As creators and IP holders argue with generative AI companies over the correct protocol to use data for training generative AI systems, a new non-profit firm, Fairly Trained, is offering certifications to companies who train their generative AI models on “consented” data. “We believe consumers deserve to know which companies think creator consent is important and which don’t. So, we certify AI companies that don’t use any copyrighted work without a license,” the company, dubbed Fairly Trained, said on its homepage. “Fairly Trained exists to make it clear which companies take a more consent-based approach to training, and are therefore…

The problems: LLM hallucinations and limited context LLMs often take a long time using expensive resources to train, sometimes months of run time using dozens of state-of-the-art server GPUs such as NVIDIA H100s. Keeping the LLMs completely up-to-date by retraining from scratch is a non-starter, although the less-expensive process of fine-tuning the base model on newer data can help. Fine-tuning sometimes has its drawbacks, however, as it can reduce functionality present in the base model (such as general-purpose queries handled well in Llama) when adding new functionality by fine-tuning (such as code generation added to Code Llama). What happens if…