Welcome to BIGDATA AI News Briefs. This timely new feature provides the latest industry insights and perspectives surrounding AI areas such as deep learning, large-scale language models, generative AI, transformers, and more. We work tirelessly to unearth the most timely and interesting information underlying the day's most popular technologies. We recognize that this field is rapidly evolving and want to provide regular resources to keep you up to date with the latest information. enjoy!
ChatGPT's new search feature – ChatGPT can now browse the Internet and provide up-to-date, authoritative information with direct links to the sources. You are no longer limited to data from before September 2021. Since first starting browsing in May, OpenAI has received valuable feedback. Updates include following robots.txt and identifying user agents that allow you to control how sites interact with his ChatGPT. Browsing is especially useful for tasks that require up-to-date information, such as assisting with technical research, choosing a bike, or planning a vacation. Browsing is currently available to Plus and Enterprise users and will soon be expanded to all users.To enable it, in the selector below GPT-4[Bing で参照]Choose.
Lessons from Meta Connect by Mike Proulx, Forrester Vice President, Research Director –
“Perhaps the star of today's Meta Connect was Meta AI, the first generation assistant based on Meta's Llama 2 language model. A partnership with Microsoft's Bing search gives you access to real-time information Emu (Meta's Expressive Media Universe image model) is also embedded in the UI, allowing users to generate custom photorealistic images and stickers. Instagram integration for advanced AI photo manipulation will be a hot topic as it brings immediate utility to the IG experience.
It was the Meta AI integration. Next generation Ray-Ban smart glasses That was probably the biggest surprise of the day. Almost (but still not quite) like the fictional Tony Stark's Jarvis-enabled glasses, his Ray-Bans allow users to interact with the Meta AI using voice commands. As Zuckerberg puts it, it's an “AI assistant that listens to what you hear and sees what you see.” Yes, this is creepy, but smart he is promising for Glass's future. ”
Cisco to buy Splunk in blockbuster $28 billion deal – Cisco just acquired analytics and cybersecurity company Splunk for $28 billion. Cisco-Splunk gains strong ARR and strong AI initiatives and capabilities, making it one of the world's largest software companies.
ChatGPT has announced that it will integrate audio and image capabilities, allowing commonly used AI tools to hear, see, and speak. Audio and images give you even more ways to use ChatGPT in your life. Take photos of landmarks during your trip and have live conversations about what's interesting about them. When you get home, take photos of your fridge and pantry to plan your dinner menu (and ask follow-up questions for step-by-step recipes). After dinner, help your child with their math problems by taking pictures, circling their problem sheets, and asking them to share tips together.
Pryon raises $100 million to index and analyze enterprise data – Pryon, a startup developing an AI-powered platform that searches for insights in enterprise knowledge bases and surfaces answers from them, today announced a funding round led by Thomas. announced that it had raised $100 million. Tal's U.S. Innovative Technology Fund.
Coqui releases model weights for new speech production model – XTTS is a speech production model that allows you to clone speech into different languages using just a short 6 second audio clip. 3 seconds cloning, multilingual voice generation, cross language voice cloning, 24khz quality.
Tabby: Open Source GitHub Copilot – Tabby is an open source, self-hosted AI coding assistant that positions itself as a viable alternative to GitHub Copilot. This makes it easy for teams to leverage Coding LLM, ensuring flexibility and compatibility for code completion tasks.
Top Author Joins Lawsuit Against OpenAI Over Novel Copyright Infringement – This lawsuit is the third by a writer to take issue with AI companies using their novels as training data without their consent.
Amazon invests $4 billion in Anthropic to compete with Microsoft (OpenAI) – Amazon plans to invest up to $4 billion in AI startup Anthropic. The two companies will work together to develop a highly reliable and high-performance platform model. Amazon Web Services (AWS) will be Anthropic's primary cloud provider. The investment also includes AWS' Trainium and Inferentia chips used to train and deploy models. This investment marks a significant move in the AI industry and positions Amazon to compete directly with other technology giants such as Microsoft, Google, and Nvidia. This collaboration aims to make AI more secure, reliable, and widely accessible. AWS customers will also have access to advanced AI capabilities.
Amazon Invests Billions of Dollars in AI Startup Anthropic – Insights from Forrester Senior Analyst Rowan Curran –
“This announcement from Amazon and Anthropic further brings together a powerful combination of enterprise expertise and advanced LLM capabilities. It also demonstrates how some of the world's largest and most technologically advanced companies can Amazon and Anthropic's partnership is similar to other announcements of this kind we've seen before. By providing compute and storage for research and development of advanced models, Anthropic provides AWS customers with advanced access to future models.
At Anthropic, one of our core principles is to create and deliver AI technology responsibly. Combined with Amazon's focus on privacy and security aspects in using its proprietary model with Bedrock, this makes it attractive for companies and governments wanting to deploy their LLM at scale in a secure cloud environment. It could be an alternative. ”
“Light Touch” AI Bill – Senate Minority Whip John Thune (R.S.D.) and Sen. Amy Klobuchar (D-Minn.) would require AI companies to self-certify the safety of high-risk AI systems. We plan to introduce a bill to make it mandatory.
AI Regulation: Be Careful and Greatly Tactful – Spending Years Building the “Perfect AI Law” and “Completely New AI Institutions” Will End Up Like the Fable of the Blind Man and the Elephant , creating repackaged laws that seek to regulate AI. From different angles without solving anything overall. We need default-acting AI regulation that balances opening the door to innovation and creativity while shadowing the misuse of data and punishing discriminatory, abusive, and biased behavior. That's what I mean.
“We don’t need the cloud” – New Intel chips scheduled for release in December will be able to run generative artificial intelligence chatbots on laptops, rather than relying on cloud data centers to get their computing power.
Roblox acquires voice moderation startup Speechly – The acquisition of voice API startup Speechly brings the company's advanced real-time voice chat and AI voice chat moderation services to online gaming networks.
Debunkd – Helps you fact check your ChatGPT results. Please ensure that the information you receive is accurate and reliable.
Phi-1.5 – Phi-1.5 is a 1.3 billion parameter Transformer language model trained primarily on synthetic datasets. This model achieves near-SOTA performance on tasks such as common sense reasoning and language understanding, especially among models with parameters less than his 10B.
Google’s BARD NEW insane update changes everything! – Bard, Google’s AI chatbot, has received a major update that now integrates seamlessly with Google services like Gmail and Maps. With features like real-time fact checking and support for 40 languages, Bard could surpass his ChatGPT. Explore Bard's new capabilities and how they're reshaping the conversational AI landscape.
See AI – Easily extract and monitor data from any website without any coding required. Train robots, schedule data extractions, and leverage pre-built robots for common use cases.Save time, automate tasks, and gain valuable insights
Amazon Alexa takes a leap forward with human-like generative AI – Amazon’s Alexa is introducing significant changes to both the voice and functionality of the ubiquitous voice assistant, and is bringing generative AI into the home. .
Research Paper: LongLoRA: Efficient Fine-Tuning of Large-Scale Language Models with Long Contexts – LongLoRA is a novel method for efficiently fine-tuning large-scale language models (LLMs) with long contexts while keeping computational costs low. Recent machine learning research presents an approach. . Training LLMs with long context sizes is typically very resource-intensive, requiring significant training time and GPU resources. The goal of this work is to extend the context size of pre-trained large-scale language models without incurring the huge computational load typically associated with this process.
How to fine-tune OpenAI's GPT-3.5 Turbo using Labelbox – Want to get the most out of a large language model like GPT-3.5? , it is well known that LLM depends on the data used to train or fine-tune it. Fine-tuning unlocks the full potential of large language models by efficiently specializing them to specific use cases and business needs.
Labelbox now provides end-to-end support for fine-tuning leading large-scale language models such as Google's Vertex AI and OpenAI's GPT-3.5. Labelbox makes it easy to: (i) Generate high-quality training datasets for fine-tuning. (ii) Export the dataset and start the fine-tuning job. (iii) continually evaluate model performance and identify gaps; With these new features, Labelbox provides an integrated platform for optimizing large language models and maintaining accuracy over time through continuous data iterations.
DDN Storage Delivers 700% Improvement in AI and Machine Learning for Image Segmentation and Natural Language Processing – DDN announces superior performance results for its AI storage platform in the first A Storage Benchmark published by the MLCommons Association.
Sign up for the free insideBIGDATA newsletter.
Join us on Twitter: https://twitter.com/InsideBigData1
Join us on LinkedIn: https://www.linkedin.com/company/insidebigdata/
Join us on Facebook: https://www.facebook.com/insideBIGDATANOW
