An important next step on our AI journey

Feb 06, 2023

AI is the most profound technology we are working on today. Whether it’s helping doctors detect diseases earlier or enabling people to access information in their own language, AI helps people, businesses and communities unlock their potential. And it opens up new opportunities that could significantly improve billions of lives. That’s why we re-oriented the company around AI six years ago — and why we see it as the most important way we can deliver on our mission: to organize the world’s information and make it universally accessible and useful.

Since then we’ve continued to make investments in AI across the board, and Google AI and DeepMind are advancing the state of the art. Today, the scale of the largest AI computations is doubling every six months, far outpacing Moore’s Law. At the same time, advanced generative AI and large language models are capturing the imaginations of people around the world. In fact, our Transformer research project and our field-defining paper in 2017, as well as our important advances in diffusion models, are now the basis of many of the generative AI applications you're starting to see today.

Introducing Bard

It’s a really exciting time to be working on these technologies as we translate deep research and breakthroughs into products that truly help people. That’s the journey we’ve been on with large language models. Two years ago we unveiled next-generation language and conversation capabilities powered by our Language Model for Dialogue Applications (or LaMDA for short).

We’ve been working on an experimental conversational AI service, powered by LaMDA, that we’re calling Bard. And today, we’re taking another step forward by opening it up to trusted testers ahead of making it more widely available to the public in the coming weeks.

Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence and creativity of our large language models. It draws on information from the web to provide fresh, high-quality responses. Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills.

Video format not supported

Use Bard to simplify complex topics, like explaining new discoveries from NASA’s James Webb Space Telescope to a 9-year-old.

We’re releasing it initially with our lightweight model version of LaMDA. This much smaller model requires significantly less computing power, enabling us to scale to more users, allowing for more feedback. We’ll combine external feedback with our own internal testing to make sure Bard’s responses meet a high bar for quality, safety and groundedness in real-world information. We’re excited for this phase of testing to help us continue to learn and improve Bard’s quality and speed.

Bringing the benefits of AI into our everyday products

We have a long history of using AI to improve Search for billions of people. BERT, one of our first Transformer models, was revolutionary in understanding the intricacies of human language. Two years ago, we introduced MUM, which is 1,000 times more powerful than BERT and has next-level and multi-lingual understanding of information which can pick out key moments in videos and provide critical information, including crisis support, in more languages.

Now, our newest AI technologies — like LaMDA, PaLM, Imagen and MusicLM — are building on this, creating entirely new ways to engage with information, from language and images to video and audio. We’re working to bring these latest AI advancements into our products, starting with Search.

One of the most exciting opportunities is how AI can deepen our understanding of information and turn it into useful knowledge more efficiently — making it easier for people to get to the heart of what they’re looking for and get things done. When people think of Google, they often think of turning to us for quick factual answers, like “how many keys does a piano have?” But increasingly, people are turning to Google for deeper insights and understanding — like, “is the piano or guitar easier to learn, and how much practice does each need?” Learning about a topic like this can take a lot of effort to figure out what you really need to know, and people often want to explore a diverse range of opinions or perspectives.

AI can be helpful in these moments, synthesizing insights for questions where there’s no one right answer. Soon, you’ll see AI-powered features in Search that distill complex information and multiple perspectives into easy-to-digest formats, so you can quickly understand the big picture and learn more from the web: whether that’s seeking out additional perspectives, like blogs from people who play both piano and guitar, or going deeper on a related topic, like steps to get started as a beginner. These new AI features will begin rolling out on Google Search soon.

?

When looking for insights, AI features in Search can distill information to help you see the big picture.

Helping developers innovate with AI

Beyond our own products, we think it’s important to make it easy, safe and scalable for others to benefit from these advances by building on top of our best models. Next month, we’ll start onboarding individual developers, creators and enterprises so they can try our Generative Language API, initially powered by LaMDA with a range of models to follow. Over time, we intend to create a suite of tools and APIs that will make it easy for others to build more innovative applications with AI. Having the necessary compute power to build reliable and trustworthy AI systems is also crucial to startups, and we are excited to help scale these efforts through our Google Cloud partnerships with Cohere, C3.ai and Anthropic, which was just announced last week. Stay tuned for more developer details soon.

Bold and responsible

It’s critical that we bring experiences rooted in these models to the world in a bold and responsible way. That’s why we’re committed to developing AI responsibly: In 2018, Google was one of the first companies to publish a set of AI Principles. We continue to provide education and resources for our researchers, partner with governments and external organizations to develop standards and best practices, and work with communities and experts to make AI safe and useful.

Whether it’s applying AI to radically transform our own products or making these powerful tools available to others, we’ll continue to be bold with innovation and responsible in our approach. And it’s just the beginning — more to come in all of these areas in the weeks and months ahead.?

人工智能是我們今天研究的最深刻的技術(shù)。無論是幫助醫(yī)生更早地發(fā)現(xiàn)疾病，還是讓人們能夠用自己的語言訪問信息，人工智能都有助于人們、企業(yè)和社區(qū)釋放他們的潛力。它開辟了新的機(jī)會，可以顯著改善數(shù)十億人的生活。這就是為什么我們六年前圍繞人工智能重新定位公司——以及為什么我們認(rèn)為它是我們完成使命的最重要方式：組織世界信息，使其普遍可訪問和有用。

從那時(shí)起，我們繼續(xù)全面投資人工智能，谷歌人工智能和DeepMind正在推進(jìn)最新技術(shù)。今天，最大的人工智能計(jì)算規(guī)模每六個(gè)月翻一番，遠(yuǎn)遠(yuǎn)超過摩爾定律。與此同時(shí)，先進(jìn)的生成人工智能和大型語言模型正在捕捉世界各地人們的想象力。事實(shí)上，我們的變形金剛研究項(xiàng)目和2017年的現(xiàn)場定義論文，以及我們在擴(kuò)散模型方面的重要進(jìn)展，現(xiàn)在是您今天開始看到的許多生成人工智能應(yīng)用程序的基礎(chǔ)。

吟游詩人簡介

當(dāng)我們將深入研究和突破轉(zhuǎn)化為真正幫助人們的產(chǎn)品時(shí)，這是一個(gè)非常令人興奮的時(shí)機(jī)。這就是我們使用大型語言模型的旅程。兩年前，我們推出了由對話應(yīng)用程序語言模型（簡稱LaMDA）支持的下一代語言和對話功能。

我們一直在開發(fā)一項(xiàng)由LaMDA提供支持的實(shí)驗(yàn)性對話人工智能服務(wù)，我們稱之為Bard。今天，我們又向前邁進(jìn)了一步，向值得信賴的測試人員開放，并在未來幾周內(nèi)向公眾更廣泛地提供它。

巴德試圖將世界知識的廣度與我們大型語言模型的力量、智慧和創(chuàng)造力相結(jié)合。它利用來自網(wǎng)絡(luò)的信息來提供新鮮、高質(zhì)量的回復(fù)。巴德可以成為創(chuàng)造力的出口，也是好奇心的發(fā)射臺，幫助您向9歲的孩子解釋美國宇航局詹姆斯·韋伯太空望遠(yuǎn)鏡的新發(fā)現(xiàn)，或了解更多關(guān)于足球界最佳前鋒的信息，然后進(jìn)行練習(xí)來培養(yǎng)你的技能。

視頻格式不受支持（有個(gè)小視頻，下面一句是簡介）

使用巴德簡化復(fù)雜的主題，例如向一名9歲的孩子解釋美國宇航局詹姆斯·韋伯太空望遠(yuǎn)鏡的新發(fā)現(xiàn)。

我們最初用輕量級型號的LaMDA發(fā)布它。這個(gè)小得多的模型需要的計(jì)算能力要少得多，使我們能夠擴(kuò)展到更多的用戶，允許更多的反饋。我們將把外部反饋與我們自己的內(nèi)部測試相結(jié)合，以確保Bard的回答符合現(xiàn)實(shí)世界信息的質(zhì)量、安全性和基礎(chǔ)性。我們對這一階段的測試感到興奮，以幫助我們繼續(xù)學(xué)習(xí)和提高Bard的質(zhì)量和速度。

將人工智能的好處帶入我們的日常產(chǎn)品

我們在使用人工智能來改善數(shù)十億人的搜索方面有著悠久的歷史。BERT是我們最早的變形金剛模型之一，在理解人類語言的復(fù)雜性方面具有革命性。兩年前，我們推出了MUM，它比BERT強(qiáng)大1000倍，對信息有更高級別和多語言的理解，可以挑選視頻中的關(guān)鍵時(shí)刻，并以更多語言提供關(guān)鍵信息，包括危機(jī)支持。

現(xiàn)在，我們最新的人工智能技術(shù)——如LaMDA、PaLM、Imagen和MusicLM——正在在此基礎(chǔ)上再接再厲，創(chuàng)造出全新的信息方式，從語言和圖像到視頻和音頻。我們正在努力將這些最新的人工智能進(jìn)步引入我們的產(chǎn)品，從搜索開始。

最令人興奮的機(jī)會之一是人工智能如何加深我們對信息的理解，并更有效地將其轉(zhuǎn)化為有用的知識——讓人們更容易找到他們正在尋找的核心并完成事情。當(dāng)人們想到谷歌時(shí)，他們經(jīng)常想到向我們尋求快速的事實(shí)答案，比如“一架鋼琴有多少把鑰匙？”但是，人們越來越多地轉(zhuǎn)向谷歌尋求更深入的見解和理解——比如，“鋼琴或吉他更容易學(xué)習(xí)嗎，每個(gè)人需要多少練習(xí)？”了解這樣的主題可能需要付出很多努力才能弄清楚你真正需要知道什么，人們通常想探索各種各樣的觀點(diǎn)或觀點(diǎn)。

在這些時(shí)刻，人工智能可以提供幫助，綜合對沒有正確答案的問題的見解。很快，您將在搜索中看到人工智能驅(qū)動的功能，這些功能將復(fù)雜的信息和多種視角提煉成易于消化的格式，因此您可以快速了解大局，并從網(wǎng)絡(luò)中了解更多信息：無論是尋找其他視角，例如同時(shí)彈鋼琴和吉他的人的博客，還是更深入地討論相關(guān)主題，例如初學(xué)者入門的步驟。這些新的人工智能功能將很快在谷歌搜索上推出。

在尋找見解時(shí)，搜索中的人工智能功能可以提煉信息，幫助您看到大局。

幫助開發(fā)人員通過人工智能進(jìn)行創(chuàng)新

除了我們自己的產(chǎn)品外，我們認(rèn)為通過建立在我們最好的模型之上，讓其他人從這些進(jìn)步中受益變得簡單、安全和可擴(kuò)展非常重要。下個(gè)月，我們將開始讓個(gè)人開發(fā)人員、創(chuàng)作者和企業(yè)入職，以便他們可以嘗試我們的生成語言API，該API最初由LaMDA提供支持，并遵循一系列模型。隨著時(shí)間的推移，我們打算創(chuàng)建一套工具和API，使其他人能夠輕松地使用人工智能構(gòu)建更具創(chuàng)新性的應(yīng)用程序。擁有必要的計(jì)算能力來構(gòu)建可靠和值得信賴的人工智能系統(tǒng)對初創(chuàng)企業(yè)也至關(guān)重要，我們很高興通過上周剛剛宣布的與Cohere、C3.ai和Anthropic的谷歌云伙伴關(guān)系來幫助擴(kuò)大這些努力。請盡快關(guān)注更多開發(fā)人員的詳細(xì)信息。

大膽而負(fù)責(zé)任

至關(guān)重要的是，我們要以大膽和負(fù)責(zé)任的方式將植根于這些模型的經(jīng)驗(yàn)帶到世界。這就是為什么我們致力于負(fù)責(zé)任地開發(fā)人工智能：2018年，谷歌是首批發(fā)布一套人工智能原則的公司之一。我們繼續(xù)為我們的研究人員提供教育和資源，與政府和外部組織合作制定標(biāo)準(zhǔn)和最佳實(shí)踐，并與社區(qū)和專家合作，使人工智能安全實(shí)用。

無論是應(yīng)用人工智能從根本上改變我們自己的產(chǎn)品，還是向他人提供這些強(qiáng)大的工具，我們都會繼續(xù)大膽創(chuàng)新，并在方法上負(fù)責(zé)任。這只是開始——在未來幾周和幾個(gè)月里，所有這些領(lǐng)域都會有更多。