Alin Vana

AI-generated code can behave in unexpected ways if developers don't thoroughly review it! Dynatrace CTO Bernd Greifender explains six key reasons why AI-generated code could make it more difficult to maintain the reliability and security of software over the long term

1

On a sunny Tuesday evening on 18 June, a very relaxed #SnowflakeUserGroup Meetup was hosted by Manuel Ifland & team in and outside the synvert Data Insights office in Munich. The group discussed the numerous innovations - especially in the area of #ML and #AI - that were announced and showcased at this years #SnowflakeSummit in San Francisco. How to combine Snowflake's latest #Geospatial features with newly gained insights from ML & AI for #NextGen #BusinessIntelligence or data apps was the scope of my talk with two live demos, including:  ➡️ Acquiring data from the Snowflake #Marketplace and from external storage  ➡️ Cleansing and transforming event and time series data using Snowflake's new set of #H3 SQL functions  ➡️ Training of a forecast model and predicting results with #SnowflakeML  ➡️ Using the instruction following #CortexAI #LLM Complete Function to classify customer sentiment and extract the underlying reasons for that sentiment derived from a customer reviews table  ➡️ Visualizing and exploring Discrete Global Grid #H3 data with #Streamlit in Snowflake (SiS) Unfortunately you couldn't make it to the Meetup in Munich and want to find out more? No problem, give this hands-on lab a try: Geospatial ML and AI using Snowflake Cortex - https://lnkd.in/egXJdMcG I'm already looking forward to the next Snowflake Meetup here in Munich or elsewhere in Germany: with practical examples, exchange of new ideas and best practices from real-world projects and the latest news from our data superheroes. #Snowflake #TechMeetUp #synvert #SnowflakeUserGroup

24

2 Kommentare

Last week, we saw a great and engaged participation at the ZHAW AI-driven Decision Intelligence conference. The interest for the topic is tangible. Developing Decision Intelligence capabilities is already a strategic priority in many organizations. Very likely, your competitor is already investing in the tools and methodologies necessary to translate data and AI in optimal decisions. #AI #ML #data #datascience

13

1 Kommentar

Very interesting insights emerging from "The Innovation Race: Reducing Risk and Navigating the AI Frontier for Future Success" report by Pure Storage. A couple of own thoughts: ◾ Data management emerges as the future centerpiece of AI-related budget allocations. This particular shift highlights a strategic pivoting towards the harnessing and leveraging of data i.e. the lifeblood of AI systems. Moreover, this shift reflects an in-depth understanding that robust data ecosystems are essential for deriving meaningful insights, while maintaining a transient advantage within an AI-driven world. ◾ Recruiting talent will be more related to the need for more specialized roles as AI applications will gradually become more complex and deeply integrated into mainstream business processes. ◾ Infrastructure costs showcase a projected notable decline, which suggests a maturation with the focus shifting towards system optimization and maintenance, rather than expansionary strategies. ◾ Staff training highlights then need for continuous workforce upskilling and reskilling. Integrating and adopting the latest knowledge and skills become a crucial component for maximizing investments in AI. AI's gradual mainstream adoption pushes organizations to shift their financial focus to value optimization and extraction strategies. The dynamic evolution of AI spending priorities highlights a strategic realignment towards data-centric investments. Organizations need to develop essential roadmaps to strategically allocate their AI budgets by focusing on areas that will yield the most pivotal and significant long-term benefits. Image source: The Wall Street Journal #artificialintelligence #AIspending #AIroadmap #emergingtechnologies #AIbudgets #AIstrategies

15

1 Kommentar

Watch Distinguished VP Analyst Mary Mesaglio dissect what the 2024 Gartner CEO Survey revealed. Comfort with risk taking is inherent in the CEO role — and that absolutely extends to generative AI, according to the 2024 Gartner CEO Survey. Explore the executive guide: https://gtnr.it/3VGipEZ #CEO #GenerativeAI #GenAI

1

The average turnover rate in retail is slightly above 60% according to the National Retail Federation. To foster communication, collaboration, and culture, many retailers are investing in new frontline capabilities like Microsoft Teams to increase the loyalty of their store associates and customers.   #frontlineworkers #MicrosoftTeams

2

#IBM #Think2024 Major announcements so far: - #Open source push: IBM is releasing its Granite models into open source, showcasing powerful and efficient Code LLMs that outperform larger models on industry benchmarks. - #InstructLab: In collaboration with Red Hat, IBM launched InstructLab, a unique model alignment technique to bring community contributions directly into large language models (#LLMs). Granite Models in detail: - Granite Code models: Ranging from 3B to 34B parameters, these models excel in tasks like code generation, bug fixing, and more, supporting 116 programming languages. #AI #Assistants: - New AI Assistants: Upcoming capabilities in watsonx include Code Assistant for Enterprise Java, Assistant for Z, and enhanced Code Assistant for Z to transform how businesses interact with their systems. Automation advances: - IBM concert: A new generative AI-powered tool for IT operations, integrating AI insights across applications to predict and resolve issues proactively. Ecosystem expansion: - Third-Party models and partnerships: IBM is broadening access to watsonx with third-party models and new collaborations, including the Meta Llama 3 model and strategic partnerships with Mistral AI and others. AI-Powered data products: - New offerings: IBM announced a new generative AI-powered data products like the IBM Data Product Hub and Data Gate for watsonx, enhancing data governance and optimization for AI workloads. Expanding GPU offerings: - NVIDIA GPU models: IBM is now offering GPU L4 and L40s models, supporting AI and mission-critical workloads on Red Hat Enterprise Linux AI and OpenShift AI.

63

#DellTechWorld Omdia You want to bring AI to your data. So on-premises, well Omdia sees hybrid as the direction of travel. Now talking about data and storage. PowerStore Prime is all flash, QLC arrays and a 5:1 data reduction, and is integrated in data protection.

3

The IDC CEO Survey 2024 shows that 45% of CEOs feel that technology vendors do not completely understand the downside risk potential of AI. This sentiment was echoed by about two-thirds of the CIOs surveyed. IDC define the following risk categories: • Security and data-related risks • Environmental and social impact • Governance Read more on the risks and challenges that are still being explored by the industry. #AIDownsideRisks #TechVendorUnderstanding #AIChallenges

4

AI-generated code can behave in unexpected ways if developers don't thoroughly review it! Dynatrace CTO Bernd Greifender explains six key reasons why AI-generated code could make it more difficult to maintain the reliability and security of software over the long term

2

Really looking forward to the #AWS #Manufacturing & Industrial Day in Zurich with the latest insights regarding * Technology trends in Manufacturing & Industrial industry * #GenerativeAI * #SAP in Manufacturing * #Digitalization and #SupplyChain * Product Connectivity * Operational efficiency through digitalization 👉 https://lnkd.in/eCxDdjvz

24

1 Kommentar

Cracking the code on human-machine relationships is essential to designing technology solutions, including those centered around AI. Start by understanding two key concepts: digital disinhibition and algorithmic aversion. Gartner Distinguished VP Analyst Mary Mesaglio explains in this video. Learn more about the intersection of humans and machines: https://gtnr.it/3U2oefj #ArtificialIntelligence #Machine #Digital

1

IBM's recent participation at the B7 highlighted key discussions on AI's future amid global regulatory advancements. Addressing the EU AI Act and IBM's role, IBM emphasized democratizing AI, citing hurdles in skills, data complexity, and ethics. Proposals included fostering open AI ecosystems and advocating for scalable, governance-driven solutions to enhance trust. Looking forward, IBM envisions AI's synergy with quantum computing, anticipating advancements in tackling global challenges through innovative governance strategies. Read more about IBM's vision for AI and quantum computing here IBM Gulf Software Distribution IBM Data, AI & Automation IBM Partner Plus

6

Microsoft unveiled Phi-3-vision, a cutting-edge multimodal AI model that combines language & vision capabilities, setting a new standard in AI performance... With 4.2 billion parameters, Phi-3-vision excels in tasks such as OCR, chart interpretation, and general visual reasoning, outperforming competitors like Claude 3 and Gemini 1.0 Pro in specific benchmarks. Phi-3-vision is part of the Phi-3 family, which includes Phi-3-mini, Phi-3-small, and Phi-3-medium models. These models are optimized for efficiency and cost-effectiveness, delivering high performance across various applications, from content creation to advanced coding assistance. The Phi-3 models are designed for deployment across a range of hardware environments, ensuring versatility and efficiency for developers and businesses. Additionally, Microsoft announced the general availability of Azure AI Studio, empowering developers to create complex AI applications and copilots for tasks like data analysis and project management. This expansion of AI capabilities through Azure AI services and GitHub Copilot Extensions underscores Microsoft's commitment to advancing AI technology and making it more accessible for diverse use cases. #AI #Microsoft #TechInnovation #MultimodalAI #ChangeYourPassword

1

Tackling #sustainability isn't easy. Get an inside look into how companies like Infosys, TurkTelekom and Magyar Telekom approach going green 🌱 in the #DataCenter: https://ow.ly/AtQG30sEMqh #VSPOne

32

Meet the team at Trade Ledger. In this video, they share how Microsoft Cloud for Financial Services and Azure OpenAI Service have made it possible to build "the first commercial banking AI in the world." https://lnkd.in/eFjM3zeM #MSFTAdvocate #Azure #AI

11

🚨NEWS ALERT🚨 Henkel - multinational chemical and consumer goods company - has revolutionized IT support for their 60,000 devices worldwide with the help of TeamViewer Tensor! 💻 🌍 Thanks to seamless integration with their existing IT systems and enhanced security with granular role and access management, Henkel now boasts smooth workflows, increased productivity, and satisfaction among their entire IT service department. Read the full success story here ➡️ https://lnkd.in/eyHh4aXy & reach out for more details 🤝 #TeamViewer #SuccessStory #Workflow #RemoteManagement

22