谷歌云被福雷斯特浪潮评为人工智能/机器学习平台领导者_AI阅读总结 — 包阅AI

At Google, AI research and development is core to delivering powerful AI/ML platforms to our customers. We’ve been at the forefront of AI innovation, developing groundbreaking technologies like transformers, compute-optimal training, and even our own specialized chips – the Tensor Processing Unit – optimized for AI workloads.

With over 20 years of experience integrating AI into applications like YouTube, Search, and Workspace, we learned what it takes to bring AI from the lab to production at global scale. All of this knowledge and experience is distilled into Vertex AI, our unified platform to manage the lifecycle of your AI investments. Vertex AI is helping customers around the globe like Shopify – take their AI from prototypes to production – irrespective of their use case, size, or underlying stack.

“Google has helped us in immense ways by almost leapfrogging the toilsome work it would take to bring the raw data in, store it, and have it readily available so we can use it in interesting ways for analytics, for training ML models, for taking it back to merchants and giving them more value out of it.” – Shopify

As customers look to move AI workloads from testing to production, choosing the right AI platform is critically important. It’s more than a simple technological decision; it’s a move with far-reaching business implications. We recognize the responsibility we have to guide our customers through this journey, and we’re deeply honored by the recognition and trust placed in us by our customers, users, and the wider AI community.

We appreciate Forrester’s recognition of Google as a Leader in this progressive market segment and the investments we make day in and day out. This year, we continued to invest in a few key areas:

Open and flexible approach

We understand that every business has unique needs and challenges. That’s why we’ve cultivated a large set of well-incented partners, making Vertex AI adaptable to diverse business environments. As Forrester states in the report,

“Google has also worked hard to nurture a large set of well-incented partners that is likely to help them increase adoption of Google Vertex AI.”

We’ve recently added AI21, Mistral AI’s latest family, Anthropic, and Meta’s Llama 3.1 models to Vertex AI Model Garden, and remain committed to offering the right tools for the job, providing them as a fully managed “Model-as-a-service” to eliminate development overheads.

Pioneering innovation and comprehensive tools for AI teams

We understand that AI initiatives require a collaborative approach, engaging teams with diverse skills and responsibilities. That’s why we offer a comprehensive toolkit designed for everyone from experienced data scientists to hobbyist developers. Be it recent innovations like Imagen 3 for precise generation of photorealistic images or even capabilities like Generative AI evaluation service to assess model performance across different use cases –Vertex AI equips your team with cutting-edge capabilities that span both predictive and generative AI. Our holistic approach allows you to tackle a wide range of use cases, from traditional machine learning tasks to the latest advancements in generative AI and no-code development. As noted by Forrester in the report:

“Google has strengths in development for genAI, AI infrastructure, and technology ecosystem… Reference customers appreciate the company’s stature in AI innovation, efficient scaling, and low-cost entry point.” “The company continues to outpace competitors in AI innovation, especially in genAI, and has a strong roadmap to expand tooling for multi-role AI teams.”

Enterprise-readiness

Our recent innovations have been centered around making AI not only accessible but also useful for enterprises. For example, we recently made Grounding with Google Search, generally available, with the ability to select when to use Google Search results and when to use the model’s training data – balancing quality with cost efficiency. Capabilities like this and Grounding with high-fidelity mode help you ensure your AI outputs are reliable.

As Forrester acknowledges in its report, “Google is certainly a good fit for existing Google Cloud customers, but it’s also very attractive to enterprises looking to accelerate their AI strategy. Given its differentiation, Google may be an attractive addition even to customers who have chosen other hyperscalers.”

We understand enterprises need predictable performance, and the ability to cost effectively scale their workloads. Provisioned Throughput in Vertex AI offers assurances for both capacity and price, with a 99.5% uptime SLA. Additionally Gemini 1.5 Flash’s context caching and Batch API, helps businesses handle large workloads and take advantage of our 1 million token context window, cost effectively.

It is exciting to see the pace of innovation in this market, and the benefits our customers are realizing from AI. We are honored to be recognized a Leader in the The Forrester Wave™: AI/ML Platforms, Q3 2024 report, as well as a Leader in Forrester’s corresponding research, The Forrester Wave™: AI Foundation Models for Language, Q2 2024 and The Forrester Wave™: AI Infrastructure Solutions, Q1 2024 report.