Werner Vogels

It was just yesterday that we saw GitHub announce a Copilot Workspace Tech Preview. Now Amazon comes out swinging with what looks like a pretty serious entry into the AI coding assistance space. Will be diving into this one in more depth in the coming days. #AWS #AIDev #GenAI

2

When we started AWS, we looked at the undesirable situation of developers spending 70% of their time on the undifferentiated muck of managing infrastructure (vs only 30% on what differentiated their customer experiences) and aimed to flip that equation on its head. It’s made a pretty big difference in innovation and developer productivity.   Roughly 18 years later, we’re trying to solve another 70-30-like misalignment for developers, where developers are telling us that they’re spending roughly 70% of their time on repetitive and tedious tasks and code.   Today, we’re launching the general availability of Amazon Q, the world’s most capable GenAI-powered assistant for accelerating software development and leveraging companies’ internal data—and that aims to remove much of the muck of repetitive and tedious coding and data tasks for developers and employees at large.   On the software development side, Q doesn’t just generate code, it also tests code, debugs coding conflicts, and transforms code from one form to another (today, developers can save months using Q to move from older versions of Java to newer, more secure and capable ones; in the near future, Q will help developers transform their .net code as well). Q Agents does multi-step planning and reasoning to allow developers to string together multiple requests and have Q implement them.   On the internal data side, most companies have voluminous troves of internally relevant data that resides in wikis, intranet pages, Salesforce, storage repositories like Amazon S3, and a bevy of other data stores and SaaS apps that are hard to access. It makes answering straightforward questions about company policies, products, business results, code, people, and many other topics hard and frustrating. Q makes this much simpler. You can point Q at all of your enterprise data repositories, and it’ll search all this data, summarize logically, analyze trends, and engage in dialog with customers about it.   We also introduced today a powerful new capability called Q Apps, which lets employees describe, in natural language, what apps they want to build on top of this internal data, and Q Apps will quickly generate that app. This is going to make it *so* much easier for internal teams to build useful apps from their own data.   Customers are gravitating to Q, and we already see companies like Brightcove, British Telecom, Datadog, GitLab, GoDaddy, National Australia Bank, NCS, Netsmart, Slalom, Smartsheet, Sun Life, Tata Consultancy Services, Toyota, and Wiz using Q. And, we’ve only been in beta til today.   Very excited about how Q will change what’s possible for our customers, and being a part of helping them innovate more quickly.

9,820

171 Comments

Sometimes small changes are a big deal. S3's per-account bucket limit has meant that buckets have been more of a human provisioning thing, starkly different from objects, which tend to be interacted with primarily from code. For years, customers have been asking us to increase this limit -- and in the storage innovation talk at re:Invent this year, i'd tell a bit of a story about why it was a more complicated bit of work than you might expect. With this limit increase, buckets have become a much more programmatic construct. We hear customers talking about using buckets, with these limit increases, as a way of managing data sharing with their own customers and using them to more effectively isolate data for specific jobs and teams. If larger numbers of buckets opens up a new pattern for you, I'd love to hear more about it. These launches, like object consistency, S3 conditionals, and big bucket limit increases have a tendency to fly a little below radar. But they are the ones that I actually hear the most excitement back from developers about, because they really broaden S3 as a programmatic interface to your data, and a building block for really incredible systems. I'm so impressed with the hard and careful work on the team that went in to getting this one out. https://lnkd.in/gA2cWATM

245

8 Comments

Since we launched Redshift Streaming Ingestion with Kinesis Data Streams (KDS) and Managed Streaming for Apache Kafka (MSK) we have seen great adoption with thousands of customers ingesting data at very high rates with low latency directly to Redshift Managed Storage. One common request was to extend the ingestion capabilities to other Kafka sources. Today we are extending Redshift's Streaming Ingestion to support self-managed Apache Kafka on EC2 as well as partners, such as Confluent Cloud. Go and Ingest! https://lnkd.in/gpzC-JD2 #amazon #redshift #streaming #ingestion #kafka #confluent

200

1 Comment

Fine tuning an LLM (Large Language Model) means adapting it to specific data and tasks with the goal of improving performance on those and similar tasks. In a new blog post, my colleagues describe a set of best practices for fine-tuning Anthropic Claude 3 Haiku on Amazon Bedrock: 1 - Choose a suitable use case per the guidelines in the post. 2 - Gather a clean, high-quality data set and structure it for efficient training, with an optional system prompt. 3 - Customize and run the fine-tuning job. 4 - Assess the fine-tuned model's performance against relevant metrics and against the base model. 5 - Iterate and deploy. Read the entire post at https://lnkd.in/gAT6kzHK to learn a lot more!

121

9 Comments

In this podcast, Google Cloud's Oliver Parker vice president, global generative AI GTM, and Miku Jha, director, AI/ML and generative AI partner engineering, sit down with Bob Evans to talk about how businesses can safely and efficiently scale their generative AI efforts, focusing on high-value opportunities while leveraging Google Cloud’s open platform, extensive ecosystem, and advanced AI/ML capabilities to deliver meaningful results for clients. Scaling Generative AI The Big Themes: Prioritizing cost and impact: Businesses are focused on the dual priorities of cost efficiency and impactful results when implementing generative AI. With numerous potential applications for generative AI, organizations need to evaluate which projects will deliver the highest value. This approach involves collaborating with partners to identify the opportunities that offer the most substantial benefits. Open ecosystem advantage: Google Cloud’s emphasis on openness and flexibility is a key differentiator in the AI landscape. Its commitment to open systems and platforms allows for various models and technologies, including first-party and open-source offerings. This open ecosystem approach enables businesses to leverage various AI models and infrastructure options, enhancing scalability and adaptability. End-to-end offerings: Google Cloud offers a comprehensive AI stack that supports a wide range of applications, from foundational models to advanced infrastructure. This end-to-end tool facilitates the development and scaling of sophisticated AI applications. The integration of various layers, including model capabilities, hardware options, and platform services, enables partners to deliver efficient and effective AI offerings. The Big Quote: “Having great models is one thing, but having differentiated infrastructure and a platform — all those three things come together, and I think we are unique in that sense.” More from Google Cloud: Learn more about Google Cloud and AI.

2

This is a worthwhile overview on the bets Satya Nadella has made at Microsoft. His overarching them has been that with AI, every old fight and every old bet the Microsoft has made was back on the table, is paying off. He's made some big bets, including in OpenAI. A few didn't pay off, and a few more are still up in the air. He's successfully put Alphabet Inc./Google back on it's heals. He's forced Cisco to step up their security game. It's fascinating to see how he's revitalized the company and made it relevent again while building its resiliency to weather a few failures along the way. #Microsoft #satyanadella #ai #innovation #competition #technology https://lnkd.in/e7ssR6JM

2

One does not simply waive a wand and make DevSecOps applicable to MLOps. Nice, concise summary of some of the basic versioning challenges associated with AI/ML model building, and why it cannot be treated as simply sequential or iterative. https://lnkd.in/e7MeXBN8

1

🚀 AWS re:Invent 2024: The Countdown Begins! 🎟️ Registration is LIVE. But the real question is: Are YOU in? 👉 Yes: What's your game plan? 🚫 No: What's holding you back? Drop your thoughts below! 👇 Let's turn this thread into the ultimate re:Invent preview. #AWS #reInvent2024 #Keepitsimple #cadyshack

5

AWS continues to add capabilities to Amazon Bedrock that make it the easiest way to build high quality, low latency, and cost-effective GenAI applications. Today the team announced:   - Custom Model Import, which lets customers easily bring their own models (from SageMaker and S3) to Bedrock to leverage its leading GenAI features, such as RAG, agents, Knowledge Bases (guardrails), inference deployment, and advanced customization to keep teaching and refining models.   - Model Evaluation, which lets customers quickly select predefined evaluation criteria (like accuracy and robustness) and run reports that help them understand which is the best model for them.   - Meta’s new Llama 3 models, now available on Bedrock and can support a wide range of use cases including text summarization and classification, sentiment analysis, language translation, and code generation.   - New Cohere Models, Command R and R+, are state-of-the-art FMs customers can use to build enterprise-grade generative AI applications with advanced RAG capabilities, in 10 languages, to support their global business operations. These will arrive on Bedrock in the coming weeks.   - New Amazon Titan models with Amazon Titan Image Generator (GA today) helping customers produce studio-quality images, at low cost, using natural language prompts (and applies an invisible watermark to help reduce the spread of disinformation)… along with Amazon Titan Text Embeddings V2 model, which reduces storage and compute costs, while out-performing other leading models on text embeddings.   Tens of thousands of customers are already building real-world production-ready GenAI apps and we're excited for even more customers to get started. https://lnkd.in/g4rkak47

2,848

54 Comments

Google Cloud Vertex AI is rolling out multiple updates to make AI accessible for every enterprise. Up to 85% lower input costs and 80% lower output costs, and 100+ new languages coming to Gemini 1.5. #GoogleCloud #VertexAI #Gemini

3

Google Cloud Vertex AI is rolling out multiple updates to make AI accessible for every enterprise. Up to 85% lower input costs and 80% lower output costs, and 100+ new languages coming to Gemini 1.5. #GoogleCloud #VertexAI #Gemini

10

Google Cloud Vertex AI is rolling out multiple updates to make AI accessible for every enterprise. Up to 85% lower input costs and 80% lower output costs, and 100+ new languages coming to Gemini 1.5. #GoogleCloud #VertexAI #Gemini

7

Since we launched DataSharing in Redshift, people really liked the efficient transactionally-consistent access to Redshift Managed Storage across Redshift compute environments, across accounts, even across regions. One of the main requests from the customers was to extend the read-only access to these live datasets, and support writes. Today we are excited to launch the general availability of multi-data warehouse writes in Redshift. #amazon #redshift #datasharing #multiwarehouse #writes #ga

83

2 Comments

To date, organizations that need to provide AI inference have had to run long running cloud instances, or provision hardware on-premises. Today, Google Cloud is previewing a new approach, and it’s one that could reshape the landscape of AI application deployment. The Google Cloud Run serverless offering is now integrating NVIDIA L4 GPUs, effectively enabling organizations to run serverless inference.

12

One of the most powerful capabilities that Redshift is launching today, Amazon Redshift now supports incremental refresh for Materialized Views (MVs) on data lake tables. With IVM on Data Lake MVs users can now combine the strengths of data lake and high performance of warehouses.

187

8 Comments

Google has big plans for AI agents. At Google Cloud Next ‘24 , Google introduced the Vertex AI Agent Builder, and previewed six AI agents they envision transforming the future of work. Google defines agents as, “Intelligent entities that take actions to achieve goals,” and Google views these agents as the cornerstone of adoption and value creation in the enterprise. Here’s an overview from Google: ***Customer Agents*** Customer agents are able to listen carefully, understand your needs, and recommend the right products and services. They work seamlessly across channels including the web, mobile, and point of sale, and can be integrated into product experiences with voice and video. ***Employee Agents*** Employee agents help workers be more productive and collaborate better together. These agents can streamline processes, manage repetitive tasks, answer employee questions, as well as edit and translate critical communications. ***Creative Agents*** Creative agents can expand your organization with the best design and production skills, working across images, slides, and exploring concepts with workers. With creative agents, anyone can become a designer, artist, or producer. ***Data Agents*** Data agents are like having knowledgeable data analysts and researchers at your fingertips. They can help answer questions about internal and external sources, synthesize research, develop new models, and help find the questions we haven’t even thought to ask yet, and then help get the answers. ***Code Agents*** Code agents are helping developers and product teams to design, create, and operate applications faster and better, and to ramp up on new languages and code bases. ***Security Agents*** Security agents assist security operations by radically increasing the speed of investigations, automating monitoring and response for greater vigilance and compliance controls. They can also help guard data and models from cyberattacks, such as malicious prompt injection. This week’s podcast covers all the Google news, plus hot AI tech (Udio and Captions), an updated GPT model from OpenAI, Adobe training data issues, and some AI legislation updates. 00:02:25 — Google Cloud Next 00:18:43 — Google’s Blockbuster Ad Deal with WPP 00:30:39 — Hot AI Tech Getting Attention - Humane AI Pin’s Brutal Early Feedback 00:37:05 — Hot AI Tech Getting Attention - Udio AI Music Generator 00:41:40 — Hot AI Tech Getting Attention - Captions AI-Powered Creative Studio 00:47:16 — Brands Adding AI Restrictions to Agency Contracts 00:51:35 — AI Legislation Updates 00:55:09 — Google’s Demis Hassabis Chafes Under New AI Push 00:58:27 — Gemini for Google Workspace Prompt Guide 01:02:23 — TikTok Plots Using Virtual Influencers for Advertising 01:04:50 — ChatGPT Is Now “More Direct, Less Verbose” 01:07:40 — Adobe’s ‘Ethical’ Firefly AI Was Trained on Midjourney Images   https://lnkd.in/gVH_V7ZF

56

9 Comments