Author: IBL News

  • Elon Musk Open-Sources Grok, But Without Training Code

    Elon Musk Open-Sources Grok, But Without Training Code

    IBL News | New York

    Elon Musk’s xAI.com was finally released as open-source under the Apache 2.0 license — which permits commercial use — the base code of its LLM ‘Grok-1’ yesterday.

    Grok-1 was released without any training code; only with the base model weights and network architecture.

    According to its description, it’s “a 314 billion parameter Mixture-of-Expert model trained from scratch by xAI”, and “not fine-tuned for any specific application, such as dialogue.”

    This release doesn’t include connections to the X social network.

    Other large companies such as Meta and Google have also released their open-source versions. LLaMA, from Meta; Gemma2B, and Gemma7B, from Google; along with Mistral, Falcon, and A12, are some of the most popular.

    Perplexity CEO Arvind Srinivas posted on X that the company will fine-tune Grok for conversational search and make it available to Pro users.
    .

     

  • Civitai.com AI Content Marketplace Reaches 4M Registered Users

    Civitai.com AI Content Marketplace Reaches 4M Registered Users

    IBL News | New York

    Civitai.com, an online community platform where members can post their own Stable Diffusion-based AI models, reached 4 million registered users, with over 12 million unique visitors per month and 10,000 unique creators.

    “We became the de facto standard for sharing your model and various AI resources and the images that you’d made,” stated the founder of the startup in an interview with TechCrunch.

    As a result of its growth, Civitai raised a $5.1 million round in June led by Andreessen Horowitz (a16z) at a $20 million valuation.

    For now, the site is free to use — it uses Cloudflare’s R2 to keep costs down around downloads.

    The company plans to allow users to monetize their work and companies to expand their brands and products.
    .

  • Coursera Introduces a Business Guide for Acquiring Skills in Generative AI

    Coursera Introduces a Business Guide for Acquiring Skills in Generative AI

    IBL News | New York

    Coursera, which provides learning pathways to upskill users, introduced this month a business guide for acquiring skills in Generative AI. [PDF download]

    This guide, titled “The Business Leader’s Playbook to Generative AI Skills Training”, features insights from leaders at Microsoft, Vanderbilt University, and Coursera.

    “GenAI is a top priority for business leaders, as it promises to transform workforce productivity,” states the report.

    “GenAI isn’t just a learning initiative. It’s a company and culture initiative that can supercharge talent and productivity.”

    Elisa Graceffo, General Manager, Technical Content, Worldwide Learning at Microsoft, said: “AI is coming to your job; it’s not coming for your job.”

    “If we can do our work more productively and focus on higher-level tasks that drive more value, then we have happier employees and more successful teams. We can reach company goals faster,” explained Trena Minudri, VP & Chief Learning Officer at Coursera.

    “Employees might feel empowered to experiment and bring innovation to their discipline.”

    Currently, Coursera hosts more than 35 GenAI-related courses from institutions like Microsoft, Stanford, Google Cloud, IBM, AWS, and Vanderbilt University, which have accumulated more than 570,000 enrollments.

    In an effort to improve productivity, Coursera’s own employees are actively learning with the company’s GenAI Academy how to use GenAI for their roles.

    According to Coursera, 89% of executives listed AI and GenAI as one of their top three tech priorities for 2024—yet two out of three (66%) executives are ambivalent or outright dissatisfied with their organization’s progress on AI/GenAI so far, citing a lack of talent and skills (62%) as their primary reason for dissatisfaction.

    “This dissonance makes sense because only 19% of HR and L&D leaders are ranking AI as a top priority for them. Mindset shifts and individual AI upskilling need to happen—and fast—so that they can meet this major inflection point in the history of technology.”

    “At the company level, when infused into the culture at large, GenAI can unify a company. It’s something everyone in the organization can learn, at the same time, whether it’s your financial analysts, marketing reps, sales leaders, C-suite, or software engineers. Everyone can benefit from prompt engineering and other foundational GenAI skills.”

    “Move too fast, and you can hit ethical and regulatory hurdles. Move too slow, and you’ll find yourself outpaced by your competitors,” added Coursera.

  • An Autonomous AI Agent Called ‘Devin’ Plans and Executes Complex Coding Tasks

    An Autonomous AI Agent Called ‘Devin’ Plans and Executes Complex Coding Tasks

    IBL News | New York

    Cognition AI, which builds AI teammates, introduced this week an autonomous AI software engineer called Devin.

    This AI agent can independently write entire software projects from scratch based on simple text prompts.

    Devin can plan and execute complex coding tasks with hundreds of steps.

    The autonomous agent can code while learning, recall relevant context at every step, fix errors, and collaborate with users in real time.

    “With our advances in long-term reasoning and planning, Devin can plan and execute complex engineering tasks requiring thousands of decisions,” said the company.

    Cognition AI has equipped Devin with common developer tools including the shell, code editor, and browser within a sandboxed compute environment.

    This agent can report on its progress in real-time, accept feedback, and work together with the user through design choices as needed.

    In the demos shown below, Devin built complete websites and apps in under 10 minutes. It also successfully completed real gigs posted on Upwork by itself.

    On a coding benchmark, the AI agent solved 13.86% of real-world GitHub issues end-to-end, crushing the previous SOTA benchmark of 1.96%.

    Funded with a $21 million Series A led by Founders Fund, Cognition AI is dedicated to building AI teammates with capabilities far beyond today’s existing AI tools by solving reasoning.
    .

  • Elon Musk Announces that It Will Open-Source Its Chatbot Grok

    Elon Musk Announces that It Will Open-Source Its Chatbot Grok

    IBL News | New York

    Elon Musk’s AI startup xAI will open-source its chatbot Grok this week, the entrepreneur said.

    The owner of Tesla and SpaceX made this announcement days after suing OpenAI and complaining that the Microsoft-backed startup had deviated from its open-source roots.

    Available to customers paying a $16 monthly subscription, Grok includes access to “real-time” information and views undeterred by “politically correct” norms.

    The promise to imminently open-source Grok will help xAI join the list of a number of growing firms, including Meta and Mistral, that have published the codes of their chatbots to the public.

    Musk has long been a proponent of open source. Tesla has open-sourced many of its patents. “Tesla will not initiate patent lawsuits against anyone who, in good faith, wants to use our technology,” Musk said in 2014.

    X, formerly known as Twitter, also open-sourced some of its algorithms last year.

    Musk reaffirmed his criticism of the Altman-led firm Monday, saying, “OpenAI is a lie.”

    On the other hand, OpenAI lambasted Elon Musk’s allegations against it, saying in a court filing that the billionaire entrepreneur’s claims “rest on convoluted — often incoherent — factual premises.”

    The strongly worded filing is the company’s first legal response to Musk’s February lawsuit against the company, Chief Executive Officer Sam Altman, and President Greg Brockman.

  • Microsoft Plans to Launch a Copilot Chatbot for Finance

    Microsoft Plans to Launch a Copilot Chatbot for Finance

    IBL News | New York

    Microsoft announced this week that it will release a Copilot chatbot that can perform and automatize key tasks for people working in finance.

    Comparing and reconciling data taken from different systems is a recurring task in finance planning and analysis.

    This Copilot for finance will initially run a variance analysis, reconcile data in Excel, and speed up the collections process in Outlook, Charles Lamanna, a Microsoft Vice President, said in an interview with CNBC.

    This tool will draw on information stored in SAP and Microsoft Dynamics 365.

    Microsoft already has a Copilot for general-purpose industrial use in Office applications.

     

  • Inflection Introduced an Upgraded Model for Its Chatbot ‘Pi’

    Inflection Introduced an Upgraded Model for Its Chatbot ‘Pi’

    IBL News | New York

    Inflection announced an upgraded model that makes its chatbot called Pi competitive with GTP-4 and Google’s Gemini, according to this AI start-up.

    “Inflection-2.5 approaches GPT-4’s performance, but used only 40% of the amount of computing for training,” said the company. “We’ve made particular strides in areas of IQ like coding and mathematics.”

    Hosted in Azure, Inflection-2.5 is available at pi.aiiOS, Android, and a new desktop app.

    Pi also incorporates real-time web search capabilities, providing users with breaking news, current events, and up-to-date information.

    According to the AI start-up, Pi has one million daily users and six million monthly active users.

    An average conversation with Pi lasts 33 minutes, and one in ten lasts over an hour each day.

    About 60% of people who talk to Pi on any given week return the following week, and we see higher monthly stickiness than leading competitors.
    .

  • OpenAI Reinstates Sam Altman as a Board Member and Announces New Ones

    OpenAI Reinstates Sam Altman as a Board Member and Announces New Ones

    IBL News | New York

    Sam Altman, as CEO, will rejoin the OpenAI Board of Directors four months later after losing his seat and being fired, the company announced in a blog post yesterday.

    Joining alongside him, three new members will be added to the Board: former CEO of the Bill and Melinda Gates Foundation, Dr. Sue Desmond-Hellmann, ex-Sony Entertainment president Nicole Seligman, and Instacart CEO Fidji Simo. These additions will bring OpenAI’s board to seven people.

    The members of the transitory board designated after Altman’s firing in November won’t be stepping down. Salesforce co-CEO Bret Taylor (OpenAI’s current board chair), Quora CEO, Adam D’Angelo, and Larry Summers, the economist and former Harvard president, will remain in their roles on the board, as will Dee Templeton, a Microsoft-appointed board observer.

    The almost all-male board has received criticism given its lack of diversity.

    • Desmond-Hellmann, in addition to heading the Bill and Melinda Gates Foundation for six years, was previously chancellor of the University of California, San Francisco, and before that, president of product development at Genentech, where she helped develop gene-targeted cancer drugs. Desmond-Hellmann is an oncologist by training is and board-certified in both internal medicine and medical oncology.

    • Nicole Seligman, an attorney and corporate director, received national attention for her representation of Lieutenant Colonel Oliver North during the Iran–Contra hearings and President Bill Clinton during his impeachment trial. Seligman was Sony’s VC and general counsel before rising through the ranks to CEO of Sony Corporation and president of Sony Corporation of America.

    • Fidji Simo, before becoming CEO of Instacart, was head of the Facebook app at Meta and the VP overseeing Meta’s various video, games, and monetization efforts. Simo also co-founded — and is currently president of — The Metrodora Foundation, a health clinic and research institute.

    Altman’s reinstatement and the board’s expansion also followed an investigation by the law firm WilmerHale, retained by OpenAI, that unanimously concluded that “Sam and OpenAI president Greg Brockman are the right leaders for OpenAI.”

    But not all at OpenAI agree.

    This week, The New York Times described Altman as a manipulative leader who undermines the credibility of those who challenged him. Both OpenAI CTO Mira Murati and Ilya Sutskever expressed concerns about Altman’s behavior prior to his ouster last year.
    .

  • Anthropic Released the Next Generation Claude 3 Model Family

    Anthropic Released the Next Generation Claude 3 Model Family

    IBL News | New York

    San Francisco-based Anthropic announced this week three new AI models, each offering different levels of capability, speed, and cost: Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus.

    The Claude 3 family of models offers a 200K context window, with the ability to handle as high as 1 million.

    They are available to use in the chatbot claude.ai and through the Claude API.

    The company said that its most advanced model Opus outperformed their rivals — namely GPT-4 and Gemini Ultra — on most of the common evaluation benchmarks for AI systems, “including undergraduate-level expert knowledge (MMLU), graduate-level expert reasoning (GPQA), basic mathematics (GSM8K), and more.”

    “It exhibits near-human levels of comprehension and fluency on complex tasks, leading the frontier of general intelligence,” said Anthropic. [See graphic below.]

    “The Claude 3 models can power live customer chats, auto-completions, and data extraction tasks where responses must be immediate and in real-time.”

    “Haiku is the fastest and most cost-effective model on the market for its intelligence category. It can read an information and data-dense research paper on arXiv (~10k tokens) with charts and graphs in less than three seconds.”

    Anthropic has become one of OpenAI’s most formidable competitors. It was formed in 2021 by former employees of OpenAI, including Daniela Amodei and her brother Dario, who serves as its chief executive officer. Most of its customers are businesses, ranging from search engine DuckDuckGo to travel guide publisher Lonely Planet. 

     

  • OpenAI Fires Back at Elon Musk’s Lawsuit Using Emails

    OpenAI Fires Back at Elon Musk’s Lawsuit Using Emails

    IBL News | New York

    OpenAI fired back at Elon Musk’s lawsuit using the owner of Tesla’s own emails.

    In a blog post authored by OpenAI’s management team – Greg Brockman, Ilya Sutskever, John Schulman, Sam Altman, Wojciech Zaremba – the Microsoft-backed startup tried to prove that Musk backed the company’s plans to become a for-profit business, and he even insisted on raising “billions of dollars” to compete against Google.

    Open AI revealed that since its inception in 2015, it had raised less than $45 million from Musk, despite his initial commitment to provide as much as $1 billion in funding. The startup also secured more than $90 million from other donors to support its research efforts, it said.

    The San Francisco – research lab’s response followed Elon Musk suing Altman, Brockman, OpenAI, and other affiliates of the firm last week, alleging the ChatGPT-maker had breached its original contractual agreements.

    These are some of the things that OpenAI said in its blog post:

    – “Elon said we should announce an initial $1B funding commitment to OpenAI. In total, the non-profit has raised less than $45M from Elon and more than $90M from other donors.”

    – “I will cover whatever anyone else doesn’t provide,” he said.

    – “As we discussed a for-profit structure in order to further the mission, Elon wanted us to merge with Tesla, or he wanted full control. Elon left OpenAI, saying there needed to be a relevant competitor to Google/DeepMind and that he was going to do it himself. He said he’d be supportive of us finding our own path.”

    – “In late 2017, we and Elon decided the next step for the mission was to create a for-profit entity. Elon wanted majority equity, initial board control, and to be CEO. In the middle of these discussions, he withheld funding. Reid Hoffman bridged the gap to cover salaries and operations.”

    – “In early February 2018, Elon forwarded us an email suggesting that OpenAI should “attach to Tesla as its cash cow”, commenting that it was “exactly right… Tesla is the only path that could even hope to hold a candle to Google. Even then, the probability of being a counterweight to Google is small. It just isn’t zero.”

    – “We’re making our technology broadly usable in ways that empower people and improve their daily lives, including via open-source contributions.”

    – “We provide broad access to today’s most powerful AI, including a free version that hundreds of millions of people use every day. For example, Albania is using OpenAI’s tools to accelerate its EU accession by as much as 5.5 years; Digital Green is helping boost farmer income in Kenya and India by dropping the cost of agricultural extension services 100x by building on OpenAI; Lifespan, the largest healthcare provider in Rhode Island, uses GPT-4 to simplify its surgical consent forms from a college reading level to a 6th grade one; Iceland is using GPT-4 to preserve the Icelandic language.”

    – “Elon understood the mission did not imply open-sourcing AGI. As Ilya told Elon: “As we get closer to building AI, it will make sense to start being less open.  The Open in openAI means that everyone should benefit from the fruits of AI after it’s built, but it’s totally OK to not share the science…”, to which Elon replied: “Yup”. 

    – “We’re sad that it’s come to this with someone whom we’ve deeply admired—someone who inspired us to aim higher, then told us we would fail, started a competitor, and then sued us when we started making meaningful progress towards OpenAI’s mission without him.”