artificialintelligence
By Matt Asay “AI models currently shine at helping so-so coders get more stuff done that works in the time they have,” argues engineer David Showalter. But is that right? Showalter was responding to Santiago Valdarrama’s contention that large language models (LLMs) are untrustworthy coding assistants. Valdarrama says, “Until LLMs give us the same guarantees [as programming languages, which consistently get computers to respond to commands], they’ll be condemned to be eternal ‘cool demos,’ useless for most serious applications.” He is correct that LLMs are decidedly inconsistent in how they res...
Info World
By Martin Heller Back in the ancient days of machine learning, before you could use large language models (LLMs) as foundations for tuned models, you essentially had to train every possible machine learning model on all of your data to find the best (or least bad) fit. By ancient, I mean prior to the seminal paper on the transformer neural network architecture, “Attention is all you need,” in 2017. Yes, most of us continued to blindly train every possible machine learning model for years after that. It was because only hyper-scalers and venture-funded AI companies had access to enough GPUs or ...
Info World
By Paul Krill Red Hat is extending its Lightspeed generative AI technology to work with the company’s Red Hat OpenShift hybrid cloud application platform as well as with Red Hat Enterprise Linux (RHEL). Announced May 7, Red Hat OpenShift Lightspeed and Red Hat Enterprise Linux Lightspeed will offer intelligent, natural language processing capabilities, intended to make OpenShift and RHEL easier for novices to use and more efficient for experienced professionals, Red Hat said. Red Hat OpenShift Lightspeed is slated for availability in late 2024. Red Hat Enterprise Linux is in the planning stage...
Info World
By Paul Krill In a bid to “deepen the public conversation about how AI models should behave,” AI company OpenAI has introduced Model Spec, a document that shares the company’s approach to shaping desired model behavior. Model Spec, now in a first draft, was introduced May 8. The document specifies OpenAI’s approach to shaping desired model behavior and how the company evaluates trade-offs when conflicts arise. The approach includes objectives, rules, and default behaviors that will guide OpenAI’s researchers and AI trainers who work on reinforcement learning from human feedback (RLHF). The com...
Info World
By Simon Bisson Both extremely promising and extremely risky, generative AI has distinct failure modes that we need to defend against to protect our users and our code. We’ve all seen the news, where chatbots are encouraged to be insulting or racist, or large language models (LLMs) are exploited for malicious purposes, and where outputs are at best fanciful and at worst dangerous. None of this is particularly surprising. It’s possible to craft complex prompts that force undesired outputs, pushing the input window past the guidelines and guardrails we’re using. At the same time, we can see outp...
Info World
By Paul Krill Red Hat has launched Red Hat Enterprise Linux AI (RHEL AI), described as a foundation model platform that allows users to more seamlessly develop and deploy generative AI models. Announced May 7 and available now as a developer preview, RHEL AI includes the Granite family of open-source large language models (LLMs) from IBM, InstructLab model alignment tools based on the LAB (Large-Scale Alignment for Chatbots) methodology, and a community-driven approach to model development through the InstructLab project, Red Hat said. The entire solution is packaged as a bootable RHEL image f...
Info World
By Paul Krill Oracle has announced Oracle Code Assist, an AI-powered coding assistant that will provide developers with context-specific suggestions that can be tailored to an organization’s best practices and codebases. Oracle did not provide an availability date for Oracle Code Assist, but said that developers at Oracle have been using the tool to build Oracle products and services. Oracle Code Assist can be used to write, upgrade, and refactor code written in most modern programming languages, the company said. Powered by large language models (LLMs) running on Oracle Cloud Infrastructure (...
Info World
By Asa Whillock The generative AI party is still raging. This zeitgeist has rocked the business world daily in a million ways, and the ground is still shifting. Now, four months into 2024, we’re starting to see businesses, particularly those with rarified pragmatic brands, starting to demand evidence of value, of the path to the true ROI derived from AI. As pragmatic voices for value rise, how do thoughtful business leaders respond? Alteryx studied exactly this question. What are the concrete pathways to AI value? We surveyed leading CIOs and board members and found a brightly lit approach to ...
Info World
By Kjell Carlsson Responsible AI (RAI) is needed now more than ever. It is the key to driving everything from trust and adoption, to managing LLM hallucinations and eliminating toxic generative AI content. With effective RAI, companies can innovate faster, transform more parts of the business, comply with future AI regulation, and prevent fines, reputational damage, and competitive stagnation. Unfortunately, confusion reigns as to what RAI actually is, what it delivers, and how to achieve it, with potentially catastrophic effects. Done poorly, RAI initiatives stymie innovation, creating hurdle...
Info World
By Anirban Ghoshal MongoDB has introduced the MongoDB AI Applications Program, or MAAP, to help enterprises jumpstart their development of generative AI applications. MAAP, according to the company, will offer enterprises strategic advice, professional services, and an integrated end-to-end technology stack from MongoDB and its partners. These partners include consultancies, foundation model (FM) providers, cloud infrastructure providers, and generative AI framework and model hosting providers including Anthropic, Anyscale, Amazon Web Services (AWS), Cohere, Credal.ai, Fireworks.ai, Google Clo...
Info World
閲覧を続けるには、ノアドット株式会社が「プライバシーポリシー」に定める「アクセスデータ」を取得することを含む「nor.利用規約」に同意する必要があります。
「これは何?」という方はこちら