Bard AI expands globally, introduces new features

AI/ML

Bard AI expands globally, introduces new features

• Available globally in 40+ languages, including Arabic, Chinese, German, Hindi, and Spanish.
• New features include listening to responses, adjustable response styles, pinned and renamed conversations, code export, shareable chat links, and image prompts.
• All updates targeted at enhancing user experience and productivity. UPDATE

Custom Document Splitter now GA in Document AI

• Document AI Workbench's Custom Document Splitter (CDS) automates splitting, classifying documents within a file.
• Helps in document validation, selection for downstream processes; reduces processing time and cost.
• Follows process of creating, training, evaluating models, API endpoint call for classification. UPDATE

Google Cloud expands generative AI model availability

• Google Cloud announces general availability of four foundation models for Vertex AI: Imagen, PaLM 2 for Chat, Codey, Chirp.
• Enables image generation, editing, chatbot support, multilingual Speech AI; Multimodal Embeddings API in preview.
• Aids in various tasks such as image classification, content recommendations, visual search. UPDATE

Llama 2 models now in Amazon SageMaker JumpStart

• Meta's Llama 2 models now accessible in Amazon SageMaker JumpStart.
• These optimized transformer models come in 7B, 13B, and 70B parameter sizes with pretrained and fine-tuned variations.
• Designed for commercial and research use, with deployment in AWS secure environments. UPDATE

Trends

Generative AI tops emerging technologies, says Forrester

• Generative AI, poised to deliver significant ROI in 2-4 years, leads Forrester's Top 10 Emerging Technologies list.
• Followed by Conversational AI, Autonomous Workplace Assistants, Decentralized Digital Identity, and Edge Intelligence.
• Emerging technologies such as XR, Web3, Zero Trust Edge will take at least five years to offer tangible value. UPDATE

Future cloud: abstracted, intelligent, composable: Forrester

• Future of cloud involves abstraction, composability, and intelligence.
• Enterprises seek cloud-neutral abstractions and data-driven automation.
• Cloud providers focus on industry-specific services, AI-driven automation, and low-code/no-code development. UPDATE

Generative AI users indifferent to data use: Forrester report

• US Generative AI usage rises to 22% with many users indifferent to data use.
• Over half are "fine" with using generative AI for company data analysis.
• Nearly half believe generative AI boosts productivity, yet feel like they're "cheating". UPDATE

App Building

Headline: AWS Lambda, EventBridge Pipes get enhanced filtering

• AWS Lambda and Amazon EventBridge Pipes support additional filtering capabilities, including suffix filtering, equals-ignore-case, and OR matching.
• Increased bounds for numeric values.
• Helps lower costs by filtering out unnecessary data, reducing the need for custom filters. UPDATE

Azure Stream Analytics Portal gets no-code editor

Azure Stream Analytics Portal introduces no-code editor for intuitive, code-free Stream Analytics job development using drag-and-drop functionality. UPDATE

Autoscale for Stream Analytics jobs now GA

Azure Stream Analytics jobs now support native autoscaling, optimizing streaming units according to changing input load patterns. UPDATE

Amazon DocumentDB rolls out index improvements

Amazon DocumentDB introduces index improvements for faster index builds and viewing index build statuses, speeding up builds by up to 14X using parallel workers. UPDATE

AWS DMS adds support for Amazon Redshift Serverless

AWS Database Migration Service (DMS) now supports Amazon Redshift Serverless as a target endpoint, enabling secure data migration to Redshift Serverless. UPDATE

Amazon Lex unveils analytics

Amazon Lex announces the general availability of analytics, providing data-driven insights, prebuilt dashboards, and a snapshot of key metrics for improved bot performance. UPDATE

Amazon Connect Wisdom extends to chat agents

Amazon Connect Wisdom, previously available only for call agents, is now available for chat agents, providing ML-powered, real-time recommended information for quick customer issue resolution. UPDATE

Orchestration

Artifact Registry introduces cleanup policies

Google’s Artifact Registry cleanup policy defines auto-delete criteria for unused artifact versions and conditions to store artifacts indefinitely. UPDATE

Programmatic execution for Amazon EMR on EKS

Amazon EMR on EKS now supports programmatic execution of Jupyter notebooks for running interactive workloads via managed endpoints, offering added flexibility to data engineers. UPDATE

Provisioning + Runtime

Amazon FSx for NetApp ONTAP adds monitoring capabilities

• Amazon FSx for NetApp ONTAP announces access to ONTAP Event Management System (EMS) logs and packet captures for monitoring and troubleshooting.
• Allows detailed insights on events like volume changes and authentication issues.
• Available now for new file systems, and soon for existing ones in all regions where FSx for ONTAP is available. UPDATE

AWS Lambda now stops recursive loops

• AWS Lambda introduces a feature to detect and stop recursive loops in Lambda functions after 16 recursive calls.
• Helps prevent unintended usage and costs due to resource misconfiguration or code defects.
• Sends events to a Dead-Letter Queue or on-failure destination if loops exceed the limit. UPDATE