Last Week in AI #235: Google's Duet AI invades Workspace, AI21 Labs hits $1.4B valuation, OpenAI's ChatGPT goes enterprise, Driverless cars face a bumpy road
Google’s Duet AI is now available in Gmail, Docs, and more for $30 a month, OpenAI launches a ChatGPT plan for enterprise customers, AI21 Labs Raises $155M in Series C Funding; Valued at $1.4 Billion.
Top News
Google’s Duet AI is now available in Gmail, Docs, and more for $30 a month
Google has launched its Duet AI assistant across all Workspace apps, including Gmail, Drive, Slides, Docs, and more. Duet AI, announced at Google's I/O developer conference, is designed to assist users in various tasks such as converting Google Docs outlines into Slides presentations, generating images, checking grammar, and summarizing documents. The AI assistant also offers app-specific features like AI-based lighting and sound adjustments and automated meeting summaries in Google Meet, and automatic summaries of long threads in Chat. However, access to Duet comes at a cost of $30 per user for large organizations, similar to Microsoft's AI system, Copilot. Despite its capabilities, users are advised to double-check Duet's outputs due to the imperfect nature of underlying AI models.
OpenAI launches a ChatGPT plan for enterprise customers
OpenAI has launched ChatGPT Enterprise, a business-focused version of its AI chatbot app, offering enhanced privacy, data analysis capabilities, and customization options. The new offering is designed to compete with similar enterprise-oriented chatbot services like Microsoft's Bing Chat Enterprise. ChatGPT Enterprise is powered by OpenAI's flagship AI model, GPT-4, and offers an admin console for managing employee usage, integrations for single sign-on and domain verification, and a dashboard with usage statistics. It also includes unlimited access to Advanced Data Analysis, a feature that allows ChatGPT to analyze data and generate insights. OpenAI emphasizes that all conversations with ChatGPT Enterprise are encrypted and it won't train models on business data sent to the platform. The company plans to further develop ChatGPT Enterprise based on customer feedback and future plans include a business offering for smaller teams and more powerful versions of Advanced Data Analysis.
It’s a Weird Time for Driverless Cars
In San Francisco, self-driving taxis operated by Cruise and Waymo are becoming a common sight, despite numerous hiccups and safety concerns. The companies have been given permission to expand their operations, charging for rides and operating 24/7. However, incidents such as a Cruise vehicle colliding with a fire truck have raised questions about the safety and reliability of these autonomous vehicles. Despite these concerns, proponents argue that self-driving cars could significantly reduce the number of road fatalities caused by human error. The technology is still being tested and refined, with San Francisco serving as a major proving ground due to its challenging driving conditions. Critics argue that the risks and inconveniences currently posed by these vehicles outweigh the potential future benefits. The future of self-driving cars could also include applications beyond private taxi services, such as self-driving shuttle operations to supplement public transportation systems.
AI21 Labs Raises $155M in Series C Funding; Valued at $1.4 Billion
AI21 Labs, an Israel-based AI and large language models (LLMs) company, has secured $155 million in a Series C funding round, raising its valuation to $1.4 billion. The funding round saw participation from investors such as Walden Catalyst, Pitango, SCB10X, b2venture, Samsung Next, Prof. Amnon Shashua, Google, and NVIDIA. The company, which has raised a total of $283 million to date, plans to use the new funds to expand its text-based generative AI services for enterprises. Founded in 2017, AI21 Labs offers a developer platform, AI21 Studio, and a consumer product, Wordtune, which uses AI to assist with reading and writing tasks. The company's clientele includes notable names like Carrefour, Clarivate, eBay, Guesty, Monday.com, and Ubisoft.
Other News
Applications
How to Use AI to Talk to Whales—and Save Life on Earth - A professor and a group of technologists are using AI to develop a synthetic whup call to communicate with humpback whales and potentially open human ears to the dialog of the entire animal kingdom.
Alibaba launches AI model that can understand images and have more complex conversations - Alibaba has launched two new AI models, Qwen-VL and Qwen-VL-Chat, which can understand images and engage in more complex conversations, allowing users to create their own AI apps without training their own systems.
OpenAI angles to put ChatGPT in classrooms with special tutor prompts - OpenAI proposes ways to use its conversational AI agent ChatGPT in classrooms, including helping language learners, formulating test questions, and teaching critical thinking skills, while acknowledging the potential for misuse and plagiarism.
Google made a watermark for AI images that you can’t edit out - Google has developed a tool called SynthID that can watermark AI-generated images in a way that is imperceptible to humans but easily detectable by AI, aiming to address the problem of deepfakes and provide a means of identifying and detecting AI imagery.
How generative AI helped train Amazon One to recognize your palm - Generative AI was used to train Amazon One, a contactless payment system that uses palm recognition, by creating millions of synthetic images of palms to boost accuracy and speed up the training process.
Intel’s Meteor Lake CPUs will use AI to make your laptop battery last longer - Intel's upcoming Meteor Lake processors will use AI to improve power management and extend battery life in laptops, with the potential to save up to 15% more energy and improve system responsiveness by up to 35%.
VFX artists show that Hollywood can use AI to create, not exploit - VFX artists at SIGGRAPH showcase how AI is being used in film and TV production to enhance the creative process, rather than replace artists, through the use of algorithmic and generative tools that augment the work of animators and effects coordinators.
Business
AI’s Share Of US Startup Funding Doubled In 2023 - More than 1 in 4 dollars invested in American startups this year has gone to an artificial intelligence-related company, with the share of investment going to AI more than doubling compared to the previous year.
Tesla's $300 Million AI Cluster Is Going Live Today - Tesla is launching a powerful supercomputer for AI applications, which will enhance its computing capabilities for training its full self-driving technology and make it one of the world's fastest supercomputers, while also investing in developing its own supercomputer, Dojo, to meet demand for GPUs and accelerate FSD training.
Google unveils enterprise AI tools, new AI chip - Google has unveiled new artificial intelligence technology and partnerships aimed at bringing AI to large businesses, including a new version of its custom-built AI chips and an enterprise-scale tool for watermarking and identifying AI-generated images.
MapReduce, TensorFlow, Vertex: Google's bet to avoid repeating history in AI - Google is trying to avoid repeating history by releasing its industry-altering technology, as it did with MapReduce and TensorFlow, with its new bet on its Google Cloud AI infrastructure Vertex AI to avoid letting competitors run off with the spoils.
DoorDash launches AI-powered voice ordering technology for restaurants - DoorDash is launching AI-powered voice ordering technology that allows restaurants to increase sales by answering all customer calls and providing curated recommendations, addressing the issue of unanswered calls and potential revenue losses.
Nearly 40% of workers think generative AI can help with workplace communication - Workers are not opposed to the implementation of generative AI in the workplace, as a survey shows that they believe it will enhance communication and automate tasks, helping them stay in their jobs longer.
UK startup Jitty raises $2M to build AI-driven home search platform - UK startup Jitty has raised $2 million in funding to develop an AI-powered property search engine that aims to improve the home buying experience by using large-language models and computer vision to understand and analyze property listings, floor plans, and photos.
Walmart's corporate employees are getting a generative AI assistant while Amazon and Apple are restricting AI in the workplace - Walmart is providing its corporate employees with an AI "assistant" tool to increase productivity, in contrast to other companies like Amazon and Apple that are restricting the use of AI in the workplace.
A16z invests in text-to-image AI generator Ideogram - Andreessen Horowitz (A16z) has invested in Ideogram AI, a start-up developing state-of-the-art AI tools for text-to-image generation, entering a competitive market dominated by AI models like DALL-E and Midjourney.
Samsung Announces Global Launch of Samsung Food, an AI-Powered, Personalized Food and Recipe Service - Samsung has announced the global launch of Samsung Food, an AI-powered food and recipe platform that offers personalized recipes, meal planning intelligence, and smart home compatibility.
We Analyzed Millions of ChatGPT User Sessions: Visits are Down 29% since May, Programming Assistance is 30% of Use - Visits to ChatGPT have declined by 29% since May, with programming assistance being the largest use-case, according to an analysis of user sessions.
China lets Baidu, others launch ChatGPT-like bots to public, tech shares jump - Chinese tech firms, including Baidu and SenseTime, have launched their AI chatbots to the public after receiving government approval, as China aims to expand the use of such products in competition with the United States.
Google introduces generative AI to Search in India, Japan - Google has introduced generative AI to its Search tool in India and Japan, allowing users to receive text or visual results to prompts, including summaries, with the option to opt in for the feature.
Poe’s new desktop app lets you use all the AI chatbots in one place - Poe, the AI chatbot platform created by Quora, has released a new Mac app and added several updates, including the ability to have multiple simultaneous conversations with the same AI bot and access to Meta's Llama 2 model, with plans for an enterprise tier in the future.
Sapeon completes $45 mn series A funding - Sapeon, a South Korean AI semiconductor company, has raised $45 million in a Series A funding round, bringing its valuation to $378 million, and plans to collaborate with investors in various sectors to offer high-quality AI services.
OpenAI is reinventing tech compensation with flat salaries, no bonuses and no negotiation as it banks on untapped potential - OpenAI is compensating its employees with a simple salary-plus-equity model, offering a flat base salary of around $300,000 and a yearly grant of approximately $500,000 in profit participation units (PPUs), making it unique in Silicon Valley.
Microsoft officially adds Bing AI chatbot to Google Chrome - Microsoft has made its Bing AI chatbot available in the Google Chrome browser, allowing users to access Bing Chat tools and seek out information or generate content.
OpenAI's business model starts to take shape - OpenAI is developing a tiered pricing structure for its AI models, with different performance levels and accessibility options, including a new Enterprise tier, to cater to various use cases and customer needs.
Concerns
‘I hope I’m wrong’: the co-founder of DeepMind on how AI threatens to reshape life as we know it - The co-founder of DeepMind discusses the potential impact of AI on mental health care and the dangers of AI in his new book, The Coming Wave, which outlines the need for containment strategies to prevent catastrophe.
Focus: Tesla braces for its first trial involving Autopilot fatality - Tesla is set to defend itself in two upcoming trials against allegations that its Autopilot feature led to fatal accidents, which will test CEO Elon Musk's claims about the technology and have significant implications for the company's future.
Elon Musk’s FSD v12 demo includes a near miss at a red light and doxxing Mark Zuckerberg - Elon Musk's FSD v12 demo includes a near miss at a red light and doxxing Mark Zuckerberg, violating Tesla's rules for its advanced driver-assist technology.
ChatGPT says no political targeting. It’s easy to break the rules. - OpenAI's ChatGPT, an AI-powered chatbot, can be used to generate tailored political messages, potentially enabling the spread of tailored disinformation in the 2024 election, despite OpenAI's ban on political campaigns using the tool.
Cruise's collisions and adapting to AI - The article discusses the recent collision between a Cruise self-driving taxi and an emergency vehicle, highlighting the challenges and complexities of integrating autonomous vehicles into existing infrastructure and societal norms.
Applying to College? Here’s How A.I. Tools Might Hurt, or Help. - The use of AI tools like ChatGPT in college applications may reshape the process, potentially impacting the way admissions officers evaluate students' unique voices and ideas.
Growing public concern about the role of artificial intelligence in daily life - A new survey by Pew Research Center reveals that a growing number of Americans are concerned about the role of artificial intelligence in daily life, with 52% expressing more concern than excitement.
Beware of the robot dog next door? - Boston Dynamics's robot dog, Spot, is being used for inspections at various industrial sites, raising questions about how humans will react to robots in everyday life, prompting researchers at The University of Texas at Austin to study human-robot interactions and societal integration.
Research
OpenAI’s Moonshot: Solving the AI Alignment Problem - OpenAI is dedicating a significant amount of computing power to solve the challenging problem of AI alignment by 2027.
An A.I. model faced off against 24 graduate students in a test of creativity—here’s how it scored - AI models like GPT-4 are capable of producing ideas that are unexpected, novel, and unique, marking one of the first examples of AI meeting or exceeding human ability for original thinking, according to a recent study.
Graph-Based Prompting and Reasoning with Language Models - Advanced prompting techniques that model the reasoning process as a graph structure, rather than a linear chain or tree, can improve the problem-solving capabilities of large language models.
AI Cracks the Code on Odor Perception - Researchers have used machine learning to develop an AI model that can describe how certain chemicals will smell, achieving human-level skill and closing a critical gap in the scientific understanding of olfaction.
Qwen-VL: A Frontier Large Vision-Language Model with Versatile Abilities - Qwen-VL is a versatile large vision-language model that allows users to upload images, audio, and videos for analysis.
Partnership awarded £1.6m to advance the manufacture of therapeutics with machine learning - Intellegens has secured a £1.6m grant from Innovate UK to apply machine learning in the field of oligonucleotide therapies, aiming to improve manufacturing productivity and yields through performance prediction and optimization of quality control strategies.
Higher Ed’s Apprehension of AI Overstated, Cengage Survey Finds - A survey conducted by Cengage reveals that while there is some apprehension, a majority of higher education instructors believe that Generative AI (GenAI) will play an increasingly important role in education, with many planning to use it to automate administrative tasks and enhance teaching.
UAE launches Arabic large language model in Gulf push into generative AI - The UAE has launched an Arabic large language model as part of its push into generative AI, aiming to gain a global perspective on the US and beyond with curated news and analysis from journalists covering politics, business, innovation, trends, and more.
Equitable Legal Remedies and the Existential Threat to Generative AI - Generative AI programs like ChatGPT and Midjourney face legal challenges regarding copyright infringement and the potential for remedies that could destroy these programs.
Websites That Have Blocked OpenAI’s GPTBot - 1000 Website Study - The study analyzes the top 1000 websites in the world to identify which sites are already blocking OpenAI's GPTBot and how they are responding.
Analysis
The A.I. Revolution Is Coming. But Not as Fast as Some People Think. - Generative A.I. has the potential to greatly increase productivity and boost the global economy, but history shows that there is typically a significant delay between the arrival of new technology and its widespread adoption.
Will Large Language Models replace Data Scientists? - Large Language Models (LLMs) have the potential to reshape the role of Data Scientists, but they are unlikely to replace them entirely, as human expertise and creativity are still necessary in tasks that LLMs cannot fully handle.
Large language models aren’t people. Let’s stop testing them like they were. - Large language models like GPT-3 and GPT-4 are being hailed for their ability to pass human tests and assessments, but researchers argue that the current evaluation techniques create an illusion of greater capabilities and call for more rigorous and exhaustive evaluation methods.
Survey finds relatively few Americans actually use (or fear) ChatGPT - Despite its prominence in the tech world, a recent survey by Pew Research found that only about 18% of Americans have actually used ChatGPT, with more people using it for entertainment or learning purposes rather than for work, and while concerns about AI in general are increasing, only a small percentage of employed individuals believe that ChatGPT will have a major impact on their job.
ChatGPT Would Vote Democrat, New Study Finds—But It's Full of Flaws - A new study claims that the chatbot ChatGPT exhibits a strong political bias towards the left, which poses a threat to democracy, but the study has been criticized for methodological flaws and providing little evidence of the bias.
Policy
Schumer to host AI forum with CEOs including Elon Musk and Mark Zuckerberg - Senate Majority Leader Charles E. Schumer will host an AI forum with top tech executives, including Elon Musk and Mark Zuckerberg, to gather input on how Congress should craft AI legislation.
Spain launches AI regulation agency in bid to become industry leader - Spain has established the Spanish Agency for the Supervision of Artificial Intelligence (AESIA) in order to regulate AI and position itself as a leader in the industry.
How do we govern AGI? - The article discusses the governance of Artificial General Intelligence (AGI) and explores three key areas of governance: evaluations, access, and international governance.
US should use chip leadership to enforce AI standards, says Mustafa Suleyman - Mustafa Suleyman suggests that the US should leverage its chip leadership to enforce AI standards.
US curbs AI chip exports from Nvidia and AMD to some Middle East countries - The US has expanded restrictions on the export of AI chips from Nvidia and AMD to include some Middle East countries, following similar restrictions imposed on China last year, although the specific risks posed by exports to the Middle East are unclear.
ABA taps prominent lawyers to tackle AI risks, opportunities - The American Bar Association is forming a task force of lawyers, including former U.S. Homeland Security Secretary Michael Chertoff, to assess the impact of AI on the legal profession and address ethical questions.
US restricts exports of some Nvidia chips to certain Middle East countries - The US has expanded restrictions on the export of Nvidia and AMD artificial intelligence chips to include certain Middle Eastern countries, although the companies have stated that the restrictions will not have an immediate material impact on their results.
X’s privacy policy confirms it will use public data to train AI models - X's updated privacy policy reveals that it plans to use the information it collects, along with publicly available data, to train its machine learning and AI models, potentially for use in Elon Musk's other AI company, xAI.
Ban or Embrace? Colleges Wrestle With A.I.-Generated Admissions Essays. - Colleges are grappling with the decision of whether to ban or embrace AI-generated admissions essays as they consider the impact of these tools on the application process.
A New Contract for Artists in the Age of Generative AI - Generative AI has raised concerns among artists and creatives who feel that their work is being used without consent or attribution, leading to a need for a new framework that protects artists' data rights and copyrights while ensuring consent and accountability in the use of AI-generated content.
OpenAI disputes authors’ claims that every ChatGPT response is a derivative work - OpenAI has responded to class-action lawsuits from book authors who claimed that ChatGPT was illegally trained on pirated copies of their books, asking the court to dismiss most of the claims and arguing that the authors misunderstand the scope of copyright and fair use.
How Do the White House’s A.I. Commitments Stack Up? - The White House has secured "voluntary commitments" from seven leading AI companies to manage the risks of artificial intelligence, but it remains uncertain how these commitments will impact the operations of these companies.
Stephen King won't forbid AI from training on his writing, and he's not afraid of AI ... yet - Stephen King believes that AI still has a lot to learn before it can successfully mimic human creativity, but he is not opposed to his works being used to teach AI about creativity.
Explainers
AI Glossary: Basic Terms All ChatGPT Users Should Know - ChatGPT, the AI-chatbot from OpenAI, is just one part of the AI landscape, and this glossary provides important terms to know about AI, including AGI, AI ethics, algorithm, bias, chatbot, deep learning, generative AI, and more.
Copyright © 2023 Skynet Today, All rights reserved.