Google's TPUs could end up costing it a billion-plus, thanks to this patent challenge
theregister.com: [Transcription] [Excerpt] Singular alleges AI accelerators built from ripped-off blueprints.
Singular's legal spat with Google dates back to late 2019 when Bates filed a lawsuit in a Massachusetts federal court against the cloud titan [PDF]. According to the complaint, Bates disclosed various technologies he had come up with to Google under a non-disclosure agreement on three occasions between 2010 and 2014. During this time, Singular said, Bates made Google aware the technologies in question were patent protected.
blog.google: Today’s investment in our Waltham Cross data centre is the latest in a series of investments that support Brits and the wider economy at large and is further evidence of Google’s continued commitment to the UK — a key country for our business and a pioneering world leader in AI, technology and science. This announcement builds on our investment in our offices with the $1bn purchase of our Central Saint Giles office in 2022, the 1 million sq ft. development in King's Cross, and the launch of our Accessibility Discovery Centre — which aims to spur the creation of accessible tech for the whole of the UK.
Samsung and Google Cloud Join Forces to Bring Generative AI to Samsung Galaxy S24 Series
googlecloudpresscorner.com: [Transcription] [Excerpt] As part of this partnership, Samsung is also one of the first customers to test Gemini Ultra, Google's most capable and largest model for highly-complex tasks. The S24 series will also use Gemini Nano, an on-device LLM delivered as part of the Android 14 operating system, the most efficient model of Gemini for on-device tasks.
A first look at the plans for our new Chicago office
blog.google: [Transcription] [Excerpt] See how we’ll update the iconic Thompson Center building in Chicago’s Loop to be more sustainable and pay homage to the building’s original design.
Enabling next-generation AI workloads: Announcing TPU v5p and AI Hypercomputer
cloud.google.com: [Transcription] [Excerpts] Enabling AI Workloads.
Achieving both scale and speed is necessary, but not sufficient to meet the needs of modern AI/ML applications and services. The hardware and software components must come together into an integrated, easy-to-use, secure, and reliable computing system. At Google, we’ve done decades of research and development on this very problem, culminating in AI Hypercomputer, a system of technologies optimized to work in concert to enable modern AI workloads.
@GoogleColab: In collaboration with our friends at @huggingface, Colab managed runtime images now include transformers installed by default. `import transformers` is all you need. We regularly update our image, but you can always force an upgrade with `!pip install transformers --upgrade
⚡️
@GoogleColab: This is a small quality of life improvement, but also enables fun features like being able to read @huggingface datasets directly from Pandas!
🙂
☑️ #94 Nov 28, 2023
A first-of-its-kind geothermal project is now operational
blog.google: [Transcription] [Excerpt] We partnered with a clean-energy startup to develop a geothermal power project in Nevada that is now contributing carbon-free energy to the electric grid.
While we’ve started with this project in Nevada, there’s potential for geothermal to grow. We recently announced a partnership with Project InnerSpace, a leading non-profit organization dedicated to the global development of geothermal energy, to accelerate the adoption of geothermal energy.
Phase I: A global geothermal resource and prospecting map
Phase II: A global deployment sprint for first-of-a-kind geothermal projects. The Geothermal Exploration Opportunities Fund (“GeoFund”)
🙂
☑️ #93 Nov 8, 2023
Google Announces Expansion of AI Partnership with Anthropic
googlecloudpresscorner.com: [Transcription] [Excerpts] Partnership includes important new collaborations on AI safety standards, committing to the highest standards of AI security, and use of TPU v5e accelerators for AI inference
SAN FRANCISCO and SUNNYVALE, Calif., Nov. 8, 2023 /PRNewswire/ -- Google today announced an expansion of its partnership with Anthropic, committing to the highest standards of AI security and safety, and that Anthropic will leverage the latest generation Cloud TPU v5e chips for AI inference.
The two companies have worked closely since Anthropic's founding in 2021, as the AI safety and research startup has built one of the largest Google Kubernetes Engine (GKE) clusters in the industry. Anthropic also currently uses AlloyDB, Google's fully managed PostgreSQL-compatible database, to handle transactional data with high performance and reliability; and Google's popular BigQuery data warehouse to analyze vast datasets and extract insights for its employees.
Google Chrome's new "IP Protection" will hide users' IP addresses
bleepingcomputer.com: [Transcription] [Excerpts] What is Google's proposed IP Protection feature?
While IP addresses are potential vectors for tracking, they are also indispensable for critical web functionalities like routing traffic, fraud prevention, and other vital network tasks.
The "IP Protection" solution addresses this dual role by routing third-party traffic from specific domains through proxies, making users' IP addresses invisible to those domains. As the ecosystem evolves, so will IP Protection, adapting to continue safeguarding users from cross-site tracking and adding additional domains to the proxied traffic.
The feature's introduction will be in stages to accommodate regional considerations and ensure a learning curve.
In its initial approach, only the domains listed will be affected in third-party contexts, zooming in on those perceived to be tracking users.
The first phase, dubbed "Phase 0," will see Google proxying requests only to its own domains using a proprietary proxy. This will help Google test the system's infrastructure and buy more time to fine-tune the domain list.
⭕️ The First Big Antitrust Trial of the Century? The Justice Department will have the next four weeks to present its case, question witnesses and present evidence. After that, the state attorneys general will have two weeks to make a supplementary case. Google will have three weeks starting Oct. 25 to make its defense. (source)
At Google Cloud, we put your interests first. This means that when you choose to work with us, we become partners on a journey of shared innovation, shared support, and shared fate. We are committed to helping you evolve as technology advances, drawing on our depth of experience to ensure you can use the latest and best technology, while keeping you safe and protected. When it comes to the rapidly developing world of generative AI, this is imperative.
“If you are challenged on copyright grounds, we will assume responsibility for the potential legal risks involved.”
Indemnity 1: Training data
Indemnity 2: Generated output indemnity
As a part of today’s announcement, products covered include:
Duet AI (Workspace, Google Cloud)
Vertex AI Search/Conversation/Text Embedding API / Multimodal Embeddings
Visual Captioning / Visual Q&A on Vertex AI
Codey APIs
What do both indemnities together mean for our customers?
It means that you can expect Google Cloud to cover claims, like copyright infringement, made against your company, regardless of whether they stem from the generated output or Google's use of training data to create our generative AI models. By offering a two-pronged generative AI indemnity protections, we are providing balanced, practical coverage for relevant types of potential claims. By offering these indemnities on our public service terms page (see below), customers will automatically receive the benefit of these terms without needing to amend their existing agreement.
🔹Terms of services here for Google Cloud and here for Workspace.
🙂
☑️ #90 Oct 11, 2023
bigtechontrial.com: Day 19: DOJ called one of Google's first employees to examine the push and pull of Google's negotiations with Apple over the years.
blog.google: [Transcription] [Excerpts] Last month we celebrated our 25th anniversary as a company, marking decades of innovation across our products and services. Today, our growing family of Pixel devices is a culmination of our latest innovations across premium hardware, helpful software and state-of-the-art AI research.
Source: Google LLC
Pixel 8, Pixel 8 Pro, Pixel Buds Pro, Pixel Watch 2 and Tensor G3 are all part of our vision to create a consistent, helpful Google for you. We're so excited about the future of the Pixel portfolio — it feels like we're just getting started.
Creating a centralized podcast destination on YouTube Music
blog.youtube: [Transcription] [Excerpts] Looking forward to 2024, we’ll be increasing our investment in the podcast experience on YouTube Music — making it a better overall destination for fans and podcasters alike.
We plan to make Podcasts on YouTube Music globally available before the end of the year.
In the coming weeks and months, we'll gather feedback to make the migration process from Google Podcasts to YouTube Music as simple and easy as possible. Once we feel the migration tools are ready, we’ll release them along with clear guidelines on how they work.
We know this transition will take time, but these efforts will allow us to build an amazing product and a single destination that rewards creators and artists and provides fans with the best Podcasts experience. For now, nothing is changing and fans will continue to have access to YouTube, YouTube Music and Google Podcasts. We’re committed to being transparent in communicating future changes with our users and podcasters and will have more to share about this process in the coming months.
Apple exec defends the decision to make Google its default search engine on iPhones and Macs
apnews.com: [Transcription] [Excerpts] WASHINGTON (AP) — A top Apple executive defended the tech giant’s decision to make Google the default search engine on Apple iPhones and Macs, saying there was no “valid alternative.’’
Testifying in the biggest antitrust trial in a quarter century, Eddy Cue, Apple’s senior vice president of services, said Tuesday that there wasn’t “anybody as good’’ as Google at helping phone and computer users search the internet.
🔹Eddy Cue is Apple’s senior vice president of Internet Software and Services, reporting to CEO Tim Cook.
🙂
☑️ #86 Sep 25, 2023
Google experts join INFER forecasting meetup to explore the future of AI
infer-pub.com: [Transcription] [Excerpts] Earlier this month, INFER hosted a special forecasting event about the future of AI, attended by select INFER forecasters and nearly two dozen forecasting enthusiasts who work at Google, joining in a personal capacity.
This event was part of INFER’s Mission: AI Advancement – a forecasting challenge to examine the current AI-boom, the future of generative AI, and its impact on society. A handful of INFER forecasters who completed the challenge were invited to attend the private event that included a group forecasting session.
10 Wacky Moments From Day One of Google’s Antitrust Trial
gizmodo.com: [Transcription] [Excerpts] [Summarizer] According to Google, its main competitor in search, Bing, isn't flopping because of anticompetitive behavior. It’s just because Bing sucks and Google is better.
Google claims it has no idea how many people switch the default search tool on their phones and browsers.
Google pays Apple $10 billion a year not to make a search engine, the DOJ says.
Apple tried to give users choice, but Google said no.
Google hid documents and conversations from the government, the DOJ says.
You use Google because you love it, not because it’s a monopoly, according to Google.
Google’s alleged monopoly behavior started in 201
Parker Brothers, Public domain, via Wikimedia Commons
Google said the government wants you to use bad products.
Employees talked about the value of making it harder to switch to Bing.
The DOJ said the future of the internet is at stake.
Google hired the same lawyers who who prosecuted Microsoft in the ‘90s.
Google Pays $10B a Year to Maintain Monopoly, US Says
bloomberg.com: [Transcription] [Excerpts] Google Says It’s No. 1 Search Tool Because Users Prefer It to Rivals
US alleges Google spends billions to maintain search monopoly
Opening statements begin 10-week trial over alleged violations
Companies choose Alphabet Inc.’s Google as the default search engine for their browsers and smartphones because it is the best one, and not because of a lack of competition, a Google lawyer said Tuesday at the start of a high-stakes antitrust trial in Washington.
Consumers use Google “because it delivers value to them, not because they have to,” John Schmidtlein, a partner at Williams & Connolly LLP who is representing the company, said during his opening statements on the first day of the trial. “Users today have more search options and ways to access information online than ever before.”
The bigger legal threat to Google, I believe, will come next year.
That’s when the government will mount a much stronger antitrust case against the company, centered on its dominance of the digital advertising marketplace. And while AI clearly presents real business opportunities for the company, the prospect that its infinite output will overrun Google’s spam defenses means that there will be challenges as well.
Casey Newton, founder and editor of Platformer
🙂
☑️ #82 Sep 6, 2023 🎂 25th Birthday (3)
I have gotten incredible enjoyement from being along for the ride for 24+ of these years
I have gotten incredible enjoyement from being along for the ride for 24+ of these years. When I joined, we were a handful of people wedged into a small office area in downtown Palo Alto above what is now a T-Mobile store.
1/
⚡️
@JeffDean: As our search traffic grew, we were desperately trying to avoid melting every Tuesday around noon (peak traffic for the week), while simultaneously improving the service. Making the index bigger (30M->70M->200M->500M->1B pages and beyond in a short span), ...
⚡️
@JeffDean: ... updated much more often (starting with updates every month, or even every two months, if things went badly, as they seemed to every March in the early days, and bringing that down to 10 days down to a day down to minutes and seconds for frequently changing parts of the index)
⚡️
@JeffDean: We never published most of the things we did in order to accomplish this scaling. Lots of cool things here like storing data on outside of disk platters for better read bandwidth, better index compression, to then storing the entire web index in RAM (in each serving cluster!)
⚡️
@JeffDean: Much of the search system scaling work led to infrastructure creation of things like MapReduce, BigTable, Spanner, protocol buffers, and other things.
⚡️
@JeffDean: I gave a talk at WSDM 2009 that talks about some of the early search systems engineering that gives a bit of a flavor of some of the developments through multiple generations of the search system.
@JeffDean: It has been deeply gratifying to work on incredibly exciting computer science and AI problems with amazing colleagues, & to help build out a suite of 10+ products that each have more than 1B users all over the world
🙂
☑️ #81 Sep 5, 2023 🎂 25th Birthday (2)
Questions, shrugs and what comes next: A quarter century of change»
blog.google: [Transcription] [Excerpts] Editor’s note: Google will officially celebrate our 25th birthday later this month. To kick off celebrations, Google and Alphabet CEO Sundar Pichai reflects on our first quarter century, including the questions and technological advancements that led to our biggest breakthroughs and most helpful products. He also shares some thoughts on the opportunity we have with AI to do things that matter on an even bigger scale.
To everyone around the world who uses our products, our employees, and our partners:
This month, Google will celebrate our 25th birthday. It’s a huge privilege to reach this milestone, made possible by the people who use our products and challenge us to keep innovating, the hundreds of thousands of Googlers past and present who have given their talents to building those products, and our partners who believe in our mission as much as we do.
about.google: [Transcription] The Google story begins in 1995 at Stanford University. Larry Page was considering Stanford for grad school and Sergey Brin, a student there, was assigned to show him around.
Founded: 1998
Founders: Larry Page and Sergey Brin
Incorporation: September 4, 1998
Initial public offering (NASDAQ): August 19, 2004
By some accounts, they disagreed about nearly everything during that first meeting, but by the following year they struck a partnership. Working from their dorm rooms, they built a search engine that used links to determine the importance of individual pages on the World Wide Web. They called this search engine Backrub.
Soon after, Backrub was renamed Google (phew). The name was a play on the mathematical expression for the number 1 followed by 100 zeros and aptly reflected Larry and Sergey's mission “to organize the world’s information and make it universally accessible and useful.”
Over the next few years, Google caught the attention of not only the academic community, but Silicon Valley investors as well. In August 1998, Sun co-founder Andy Bechtolsheim wrote Larry and Sergey a check for $100,000, and Google Inc. was officially born. With this investment, the newly incorporated team made the upgrade from the dorms to their first office: a garage in suburban Menlo Park, California, owned by Susan Wojcicki (employee #16 and former CEO of YouTube). Clunky desktop computers, a ping pong table, and bright blue carpet set the scene for those early days and late nights. (The tradition of keeping things colorful continues to this day.)
Even in the beginning, things were unconventional: from Google’s initial server (made of Lego) to the first “Doodle” in 1998: a stick figure in the logo announcing to site visitors that the entire staff was playing hooky at the Burning Man Festival. “Don't be evil” captured the spirit of our intentionally unconventional methods. In the years that followed, the company expanded rapidly — hiring engineers, building a sales team, and introducing the first company dog, Yoshka. Google outgrew the garage and eventually moved to its current headquarters (a.k.a.“The Googleplex”) in Mountain View, California. The spirit of doing things differently made the move. So did Yoshka.
The relentless search for better answers continues to be at the core of everything we do. Today, Google makes hundreds of products used by billions of people across the globe, from YouTube and Android to Gmail and, of course, Google Search. Although we’ve ditched the Lego servers and added just a few more company dogs, our passion for building technology for everyone has stayed with us — from the dorm room, to the garage, and to this very day.
Talking not Typing: Continental and Google Cloud Equip Cars with Generative Artificial Intelligence
continental.com: [Transcription] [Excerpts] Munich, Germany, September 4, 2023. Continental announced a partnership with Google Cloud during its press conference at the IAA MOBILITY 2023. Together, the two companies are equipping cars with generative Artificial Intelligence (AI), making Continental one of the first automotive suppliers to integrate Google Cloud capabilities directly into vehicle computers. With this, drivers can interact with their car in a natural dialogue. No matter if drivers for example need help finding the right tire pressure when the car is fully loaded or want to know more about local places of interest on their destination or along their route when they are on vacation, the generative AI can compile the required information and answer the drivers’ questions. The two companies each bring their respective expertise to the partnership, namely in automotive, software, artificial intelligence and cloud computing.
Smart Cockpit High-Performance Computer: inexpensive and production-ready in just 18 months
With generative Artificial Intelligence Continental and Google Cloud enable an intuitive user experience in the vehicle. Source: Continental AG
“Together with Google, we are equipping cars with artificial intelligence for drivers. This is how our vision of software-defined vehicles starts to become a reality”
Philipp von Hirschheydt, Continental Executive Board member and head of the Automotive group sector
mimestream.com: [Transcription] [Excerpts] Today we released Mimestream 1.1, and we’re thrilled to announce that Mimestream has been approved by Google for use with the Advanced Protection Program. Support for this program has been a top request from security-conscious and enterprise users.
While we don’t recommend it for most users, if you’re very concerned about targeted and sophisticated phishing attacks, the program requires hardware security keys and introduces other measures to provide the highest level of protection to your Google account.
To get started, you will need to purchase at least two FIDO hardware security keys – one as a primary, and one as a backup to be kept in a safe location. Source: Mimestream, LLC.
cloud.google.com: [Transcription] [Excerpts] Today we are announcing key infrastructure advancements to help customers, including:
Cloud TPU v5e: Our most cost-efficient, versatile, and scalable purpose-built AI accelerator to date. Now, customers can use a single Cloud TPU platform to run both large-scale AI training and inference. Cloud TPU v5e scales to tens of thousands of chips and is optimized for efficiency. Compared to Cloud TPU v4, it provides up to a 2x improvement in training performance per dollar and up to a 2.5x improvement in inference performance per dollar.
A3 VMs with NVIDIA H100 GPU: Our A3 VMs powered by NVIDIA’s H100 GPU will be generally available next month. It is purpose-built with high-performance networking and other advances to enable today’s most demanding gen AI and large language model (LLM) innovations. This allows organizations to achieve three times better training performance over the prior-generation A2.
General Motors Teams Up with Google Cloud on AI Initiatives
investor.gm.com: Companies share new details on AI collaboration and announce future generative AI explorations
DETROIT and SUNNYVALE, Calif. – General Motors and Google Cloud today shared new details on how the two companies collaborated to bring conversational AI technology into millions of GM vehicles, helping drivers in a variety of ways. Since its launch in 2022, GM’s OnStar Interactive Virtual Assistant (IVA), has been powered by advanced, intent-recognition algorithms that use Google Cloud’s conversational AI technologies, providing OnStar Members with responses to common inquiries, as well as routing and navigation assistance. The successful deployment of Google Cloud’s AI in GM’s OnStar service has now opened the door to future generative AI deployments being jointly piloted by GM and Google Cloud.
“Generative AI has the potential to revolutionize the buying, ownership, and interaction experience inside the vehicle and beyond, enabling more opportunities to deliver new features and services,”
“Our software-led approach has accelerated the creation of compelling services for our customers while driving increased efficiency across the GM enterprise. The work with Google Cloud is another example of our efforts to transform how customers engage with our products and services.
Mike Abbott, executive vice president, Software and Services, GM.
The Path to $4.5B ✅ > Funding Round (Hugging Face)
@ClementDelangue: Super excited to welcome our new investors @SalesforceVC, @Google, @amazon, @nvidia, @AMD, @intel, @QualcommVenture, @IBM & @sound_ventures_ who all participated in @huggingface’s $235M series D at a $4.5B valuation to celebrate the crossing of 1,000,000 models, datasets and apps on the platform. These partners alone already shared over 1,000 open models and datasets and have over 10,000 users on Hugging Face. It takes a village to democratize good machine learning thanks to open-source and we’re just getting started!
The First Big Antitrust Trial of the Century Is About to Start
thebignewsletter.com: Google's monopoly is going on trial starting September 12. This is the big one. Or rather, it's the start of a series of big ones.
Review of Google’s Quantum Computing Technology State in 2023
thequantuminsider.com: [Transcription] Google is exploring quantum computing’s potential in financial services with J.P. Morgan. In this partnership, new algorithms will be developed for the optimization of portfolios, the analysis of risk and the detection of fraud.
Announcing the inaugural North American Google for Startups Accelerator: Cloud »
g.co: [Transcription] Google for Startups Accelerator: Cloud is a 10-week accelerator program for cloud-native technology startups headquartered in North America. The accelerator program is best suited for revenue generating startups with a minimum of 5+ employees. To ensure Google can provide meaningful value, startups should aim to leverage Cloud-Native technologies (including AI/ML, analytics, etc.) in their product, service, or operations and/or show an interest in leveraging these technologies in the future.
The accelerator is designed to bring the best of Google's programs, products, people and technology to startups doing interesting work in the Cloud. In addition to Cloud mentorship and technical project support, the accelerator also includes deep dives and workshops focused on product design, customer acquisition and leadership development for startup founders and leaders.
Google DeepMind: Bringing together two world-class AI teams
blog.google: [Transcription] Editor’s Note: Earlier today we announced some changes that will accelerate our progress in AI and help us develop more capable AI systems more safely and responsibly. Here is a recap of what Google and Alphabet CEO Sundar Pichai and DeepMind CEO Demis Hassabis shared with employees.
We’ve been an AI-first company since 2016, because we see AI as the most significant way to deliver on our mission. Since then, we’ve used AI to improve many of our core products, from Search, YouTube and Gmail to the incredible camera in Pixel phones. We’ve helped businesses and developers harness the power of AI via Google Cloud, and we’ve shown AI’s potential to address societal issues like health and climate change.
Along the way, we’ve been lucky to have two world-class research teams leading the entire industry forward with foundational breakthroughs that have ushered in a new era of AI.
The pace of progress is now faster than ever before. To ensure the bold and responsible development of general AI, we’re creating a unit that will help us build more capable systems more safely and responsibly.
This group, called Google DeepMind, will bring together two leading research groups in the AI field: the Brain team from Google Research, and DeepMind. Their collective accomplishments in AI over the last decade span AlphaGo, Transformers, word2vec, WaveNet, AlphaFold, sequence to sequence models, distillation, deep reinforcement learning, and distributed systems and software frameworks like TensorFlow and JAX for expressing, training and deploying large scale ML models.
Combining all this talent into one focused team, backed by the computational resources of Google, will significantly accelerate our progress in AI.
As CEO of Google DeepMind, Demis Hassabis will lead the development of our most capable and responsible general AI systems — research that will help power the next generation of our products and services.
Jeff Dean will take on the elevated role of Google’s Chief Scientist, reporting to me. In that capacity he’ll serve as Chief Scientist to Google Research and Google DeepMind. Working alongside Demis, Jeff will help set the future direction of our AI research and head up our most critical and strategic technical projects related to AI, the first of which will be a series of powerful, multimodal AI models.
Google Research will continue its important work leading fundamental advances in computer science across areas such as algorithms and theory, privacy and security, quantum computing, health, climate and sustainability and responsible AI, and will report in to James Manyika along with his existing Tech & Society teams.
I’m so excited for the next phase of this journey, the progress we’ll make against our mission, and all the ways we’ll help people reach their potential with increasingly capable and responsible AI.
blog.google: [Transcription] [Excerpt] A Google AI expert breaks it down
We recently expanded access to Bard, an early experiment that lets you collaborate with generative AI. Bard is powered by a large language model, which is a type of machine learning model that has become known for its ability to generate natural-sounding language. That’s why you often hear it described interchangeably as “generative AI.” As with any new technology, it's normal for people to have lots of questions — like what exactly generative AI even is.
WSJ News Exclusive: Google CEO Sundar Pichai Says Search to Include Chat AI
wjs.com: ‘More work left to do’ in efficiency drive, chief executive says
Google Brain and DeepMind
[Transcription] [Excerpt] Google will continue to improve Bard with new AI models, Mr. Pichai said, while declining to comment on when the product would become freely available without a wait list.
AI technology requires enormous computing power to process the calculations used to produce humanlike conversation. Mr. Pichai said Google needs to adapt its use of resources to continue its work in AI while also managing costs. For example, he said Google Brain and DeepMind—the company’s two main AI units, which have long operated separately—would work together more closely on efforts to build large algorithms.
“I expect a lot more, stronger collaboration, because some of these efforts will be more compute-intensive, so it makes sense to do it at a certain scale together,”
Top Search Engine by Country - March 2023 Market Share: Analyze the most used & most popular search engine in each country in March 2023 by market share, understand search market trends and build the right search strategy for each region. Source: Similarweb LTD
🙂
☑️ #65 Apr 3, 2023
S&P 500 Map - Year to Date performance
finviz.com: Standard and Poor's 500 index stocks categorized by sectors and industries. Size represents market cap.
Replit and Google Cloud Partner to Advance Generative AI for Software Development
blog.replit.com: Under the new partnership, Replit developers will get access to Google Cloud infrastructure, services, and foundation models via Ghostwriter, Replit's software development AI, while Google Cloud and Workspace developers will get access to Replit’s collaborative code editing platform.
[Transcription] How long it will rake before google bard will be shut down?
Google Bard is currently still under development, and there is no official announcement from Google regarding the shutdown of this project. However, based on my knowledge, it is likely that Google Bard will b shut down within the next 1-2 years.
Google Bard was launched in February 2023, and it has been me mixed reviews from users. Some users find it to be a helpful tool generating creative text formats, while others find it to be unreliable and inaccurate.
Google has cited the lack of adoption as the reason for the shutdown of Google Bard. This means that not enough users are using this tool make it a viable project for Google.
Google has also expressed its commitment to continuing to explore ways to use Al to enhance creative expression. This means that Google is likely to continue to develop new Al-powered tools for creative writing, even if it means shutting down existing projects like Google Bard.
[Transcription] We’re starting to open access to Bard, an early experiment that lets you collaborate with generative AI. We're beginning with the U.S. and the U.K., and will expand to more countries and languages over time.
Meet Bard: an early experiment by Google. Source: https://www.blog.goole
What’s Bard all about? We’ve long seen the potential to make information and computing much more accessible through conversational AI.
Two years ago, we unveiled LaMDA (Language Model for Dialogue Applications), a conversational AI model capable of fluid, multi-turn dialogue, and last year, we launched the AI Test Kitchen, a new space where people could learn about, get hands-on experience with, and provide feedback on LaMDA.
Bard is an experiment based on this same technology that lets you collaborate with generative AI. As a creative and helpful collaborator, Bard can supercharge your imagination, boost your productivity, and help you bring your ideas to life—whether you want help planning the perfect birthday party and drafting the invitation, creating a pro & con list for a big decision, or understanding really complex topics simply.
If you’re interested in the more technical details, LaMDA is a Transformer-based model, the machine-learning breakthrough invented by Google in 2017. The language model learns by “reading” trillions of words that help it pick up on patterns that make up human language so it’s good at predicting what might be reasonable responses.
x.company/glass: Thank you for over a decade of innovation and partnership. As of March 15, 2023, we will no longer sell Glass Enterprise Edition. We will continue supporting Glass Enterprise Edition until September 15, 2023.
Google discontinued Glass Enterprise Edition. Source: Google, Inc.
blog.google: At The Check Up, we shared updates on our medical LLM research and ways AI can improve maternal care, cancer treatments and tuberculosis screening.
Tune in on March 14, 2023 at 11:00am ET to hear from Dr. Karen DeSalvo, our Chief Health Officer, and other teams across Google about the exciting product launches, partnership announcements, and latest innovations from Google Health, our company-wide effort to help billions of people be healthier.
Google Cloud brings generative AI to developers, businesses, and governments
cloud.google.com: Generative AI in Google. Google brings new AI capabilities to developers & businesses
[Transcription] [Excerpt] To address these needs, Google Cloud will launch a range of products that infuse generative AI into our offerings, empowering developers to responsibly build with enterprise-level safety, security, and privacy. This journey starts today with the introduction of two new technologies:
Generative AI support in Vertex AI gives data science teams access to foundation models from Google and others, letting them build and customize atop these models on the same platform they use for homegrown ML models and MLOps.
Generative AI App Builder allows developers to quickly ship new experiences including bots, chat interfaces, custom search engines, digital assistants, and more. Developers have API access to Google’s foundation models and can use out-of-the-box templates to jumpstart the creation of gen apps in minutes or hours.
New AI partnerships and programs: In addition to announcing new Google Cloud AI products, we’re also committed to being the most open cloud provider. We’re expanding our AI ecosystem and specialized programs for technology partners, AI-focused software providers and startups.
The next generation of AI for developers and Google Workspace
@ThomasOrTK: Today marks the next chapter of our AI journey with new tools and experiences across Google Workspace and Google Cloud that allow users, developers and businesses to harness the power of generative AI.
@ThomasOrTK: We’re working with @Uber to reimagine the customer experience for users, drivers and merchants by leveraging our reliable, secure-by-design foundation combined with innovative AI and data analytics.
Tweet by @gcodingcomps: 20 years, more than a million participants and billions of lines of code later, our coding competitions are coming to a close. It's been an honor to learn and enjoy coding with you. Thank you.
@paulg: What happens if you take too long to launch: your product is defined by its relationship to whatever launched first. That casual appositive phrase is worth more to Microsoft than any news story.
paulg: Being super slow to ship is a long-established tradition at Google. I didn't say so at the time, but this essay I wrote in 2008 (15 years ago!) was inspired by the experience a YC startup had after being acquired by Google:
One of the differences between big companies and startups is that big companies tend to have developed procedures to protect themselves against mistakes. A startup walks like a toddler, bashing into things and falling over all the time. A big company is more deliberate.
The gradual accumulation of checks in an organization is a kind of learning, based on disasters that have happened to it or others like it. After giving a contract to a supplier who goes bankrupt and fails to deliver, for example, a company might require all suppliers to prove they're solvent before submitting bids.
As companies grow they invariably get more such checks, either in response to disasters they've suffered, or (probably more often) by hiring people from bigger companies who bring with them customs for protecting against new types of disasters.
It's natural for organizations to learn from mistakes. The problem is, people who propose new checks almost never consider that the check itself has a cost.
Every check has a cost. For example, consider the case of making suppliers verify their solvency. Surely that's mere prudence? But in fact it could have substantial costs. There's obviously the direct cost in time of the people on both sides who supply and check proofs of the supplier's solvency. But the real costs are the ones you never hear about: the company that would be the best supplier, but doesn't bid because they can't spare the effort to get verified. Or the company that would be the best supplier, but falls just short of the threshold for solvency—which will of course have been set on the high side, since there is no apparent cost of increasing it.
Whenever someone in an organization proposes to add a new check, they should have to explain not just the benefit but the cost. No matter how bad a job they did of analyzing it, this meta-check would at least remind everyone there had to be a cost, and send them looking for it.
If companies started doing that, they'd find some surprises. Joel Spolsky recently spoke at Y Combinator about selling software to corporate customers. He said that in most companies software costing up to about $1000 could be bought by individual managers without any additional approvals. Above that threshold, software purchases generally had to be approved by a committee. But babysitting this process was so expensive for software vendors that it didn't make sense to charge less than $50,000. Which means if you're making something you might otherwise have charged $5000 for, you have to sell it for $50,000 instead.
The purpose of the committee is presumably to ensure that the company doesn't waste money. And yet the result is that the company pays 10 times as much.
Checks on purchases will always be expensive, because the harder it is to sell something to you, the more it has to cost. And not merely linearly, either. If you're hard enough to sell to, the people who are best at making things don't want to bother. The only people who will sell to you are companies that specialize in selling to you. Then you've sunk to a whole new level of inefficiency. Market mechanisms no longer protect you, because the good suppliers are no longer in the market.
Such things happen constantly to the biggest organizations of all, governments. But checks instituted by governments can cause much worse problems than merely overpaying. Checks instituted by governments can cripple a country's whole economy. Up till about 1400, China was richer and more technologically advanced than Europe. One reason Europe pulled ahead was that the Chinese government restricted long trading voyages. So it was left to the Europeans to explore and eventually to dominate the rest of the world, including China.
In more recent times, Sarbanes-Oxley has practically destroyed the US IPO market. That wasn't the intention of the legislators who wrote it. They just wanted to add a few more checks on public companies. But they forgot to consider the cost. They forgot that companies about to go public are usually rather stretched, and that the weight of a few extra checks that might be easy for General Electric to bear are enough to prevent younger companies from being public at all.
Once you start to think about the cost of checks, you can start to ask other interesting questions. Is the cost increasing or decreasing? Is it higher in some areas than others? Where does it increase discontinuously? If large organizations started to ask questions like that, they'd learn some frightening things.
I think the cost of checks may actually be increasing. The reason is that software plays an increasingly important role in companies, and the people who write software are particularly harmed by checks.
Programmers are unlike many types of workers in that the best ones actually prefer to work hard. This doesn't seem to be the case in most types of work. When I worked in fast food, we didn't prefer the busy times. And when I used to mow lawns, I definitely didn't prefer it when the grass was long after a week of rain.
Programmers, though, like it better when they write more code. Or more precisely, when they release more code. Programmers like to make a difference. Good ones, anyway.
For good programmers, one of the best things about working for a startup is that there are few checks on releases. In true startups, there are no external checks at all. If you have an idea for a new feature in the morning, you can write it and push it to the production servers before lunch. And when you can do that, you have more ideas.
At big companies, software has to go through various approvals before it can be launched. And the cost of doing this can be enormous—in fact, discontinuous. I was talking recently to a group of three programmers whose startup had been acquired a few years before by a big company. When they'd been independent, they could release changes instantly. Now, they said, the absolute fastest they could get code released on the production servers was two weeks.
This didn't merely make them less productive. It made them hate working for the acquirer.
Here's a sign of how much programmers like to be able to work hard: these guys would have paid to be able to release code immediately, the way they used to. I asked them if they'd trade 10% of the acquisition price for the ability to release code immediately, and all three instantly said yes. Then I asked what was the maximum percentage of the acquisition price they'd trade for it. They said they didn't want to think about it, because they didn't want to know how high they'd go, but I got the impression it might be as much as half.
They'd have sacrificed hundreds of thousands of dollars, perhaps millions, just to be able to deliver more software to users. And you know what? It would have been perfectly safe to let them. In fact, the acquirer would have been better off; not only wouldn't these guys have broken anything, they'd have gotten a lot more done. So the acquirer is in fact getting worse performance at greater cost. Just like the committee approving software purchases.
And just as the greatest danger of being hard to sell to is not that you overpay but that the best suppliers won't even sell to you, the greatest danger of applying too many checks to your programmers is not that you'll make them unproductive, but that good programmers won't even want to work for you.
Steve Jobs's famous maxim "artists ship" works both ways. Artists aren't merely capable of shipping. They insist on it. So if you don't let people ship, you won't have any artists.
🙂
☑️ #52 Feb 8, 2023
@tanayj: Google lost $100B of market cap today as its Chatbot Bard made a factual error during its first-ever demo and its AI event fell flat. Must be the most costly live demo fail of all time
DataTrekMB: 1/2 #ChatGPT, the AI-powered answer bot, has the potential to change how humans interface with technology as they search for useful information. $GOOG gives you suggestions. ChatGPT gives you answers...
DataTrekMB: 2/2 ...The most important question now is “how will $MSFT and other Big Tech companies monetize AI-powered search?” Until there is a clear answer, this is just a cool new technology.
