Data Phoenix Digest - ISSUE 7.2024

Welcome to this week's edition of Data Phoenix Digest!

Be active in our community and join our Slack to discuss the latest news, events of our community, research papers, articles, jobs, and more...

Data Phoenix video library

Explore recordings of all our past webinars to deepen your AI knowledge and enhance your learning journey:

• Exploring Infrastructure Management for GenAI Beyond Kubernetes
• Evaluating LLM Models for Production Systems: Methods and Practices
• Democratizing AI Deployment
• GPT on a Leash: Evaluating LLM-based Apps & Mitigating Their Risks
• Building Customized CV Applications with FiftyOne Plugins
• Large Language Models for Program Synthesis
• Lessons Learned from Building a Managed RAG Solution
• and more

AI Highlights of the Past Week

This week's highlights are Microsoft's and Apple's tiny language models meant to be run locally in resource and network connection-constrained devices. Microsoft announced the Phi-3 family of models, with three models ranging from 3.8B to 14B parameters. The first official launch is Phi-3-mini, a 3.8 billion parameter model small enough to be run in contexts with constrained hardware and network resources, but performant enough to surpass Llama 3 8B's performance in the MMLU benchmark. Its strong performance results from a carefully curated dataset and training processes reminiscent of those reported by the team behind the Llama 3 model family.

Apple's models are minimal by comparison. The OpenELM family of models comprises four pre-trained models ranging from 270 million to 3 billion parameters, and their instruction-tuned variants. In addition to model weights and inference code, the release includes a complete training and evaluation framework that works with publicly available datasets. A key differentiator of the OpenELM models is the approach to non-uniform parameter allocation within each transformer model layer. The research paper reports that the 1.1 billion-parameter OpenELM variant achieves 2.36% higher accuracy than the 1.2 billion-parameter OLMo in OpenLLM Leaderboard tasks, while pre-trained with half as much data as OLMo. Although the OpenELM family may not be more than a proof-of-concept, it is still exciting news about a future of generative AI in which models can be run anywhere.

Two more models were recently launched, although with vastly different purposes: Snowflake launched Arctic, its dense mixture-of-experts LLM geared towards enterprise use cases; and Adobe released its next-generation Firefly Image 3 to advance its image-generation services. The model is available in beta, powering image generation and generative editing features on Photoshop and the Firefly web application. Adobe reports that Firefly Image 3 was designed to generate content appropriate for commercial use that does not violate copyright or other intellectual property rights. This characterizes a second common thread in the headline selection for the week: the variety of solutions geared specifically toward enterprises being released to the market. OpenAI introduced enterprise-grade features, updates, and cost-management tools; Perplexity AI launched its Enterprise Pro service; and Cohere launched a new toolkit to empower developers to build AI applications.

Relatedly, more industry leaders in very diverse markets are buying themselves into strategic partnerships with Microsoft/OpenAI, or simply going all in on work they started some time ago:

• United Airlines is streamlining some of its processes with the help of AI: United Airlines is taking the next step of its journey into the digital era by exploring some use cases for AI to streamline its internal operations. From its customer service chatbot to developing rerouting tools, the airline has demonstrated it is not fearful of the upcoming AI revolution.
• The Coca-Cola Company and Microsoft entered a cloud and generative AI initiative-accelerating partnership: Coca-Cola and Microsoft recently expanded their strategic partnership, with the former selecting Microsoft as its preferred and strategic partner. Currently, the partnership focuses on providing Coca-Cola with expanded access to Microsoft's cloud and AI platforms and solutions.
• The Estée Lauder Companies and Microsoft further their partnership with the AI Innovation Lab: The Estée Lauder Companies (ELC) and Microsoft recently announced their collaboration on an AI Innovation Lab. Two major developments leveraging the AI Innovation Lab are the creation of an internal-facing chatbot to boost marketing effectiveness and acceleration of product development.
• BBC is using AI to redesign its educational services: BBC recently announced a £6 million investment to transform the online study support resource BBC Bitesize. New tools developed for the platform likely include recommendation systems and personalized testing with custom ethical algorithms at their core.

While NVIDIA expands its domains, the UK's Competition and Markets Authority (CMA) wants opinions on whether the partnerships or arrangements between Amazon and Anthropic, Microsoft and Mistral AI, and Microsoft and Inflection AI fall under the jurisdiction of UK merger laws, and on the impact they may have on the Foundation Model and AI markets.

• NVIDIA is acquiring the GPU Orchestration Software Provider Run:ai: NVIDIA is acquiring Run:ai, the Kubernetes-based workload management and orchestration software provider. NVIDIA plans to offer Run:ai products under the same business model and invest in the Run:ai product roadmap.
• Markets Authority (CMA) looks for comments on key AI partnerships: The Competition and Markets Authority (CMA) opened invitations to comment (ITCs) as it gathers information on whether the arrangements between several tech firms are within the jurisdiction of UK merger rules and the effect of such partnerships on competition within the AI markets in the UK.

AI-powered testing seems to be having a bit of a moment:

• Octomind secured $4.8 million to accelerate bug-free AI-assisted software testing: Octomind recently announced it completed a successful $4.8 million seed funding round led by Cherry Ventures and backed by an outstanding selection of angel investors. The funds will enable Octomind to enhance the capabilities of its AI to automate the software testing process.
• Mabl added mobile application testing to its solution portfolio: Mabl, the leading unified test automation platform, has announced it will launch mobile application testing capabilities to its platform. Mabl's AI-powered mobile app testing promises up to 90% improvement in test coverage and reliability while reducing test creation time to as little as 2 minutes.

More headlines of interest include using AI to develop a coffee bean blend, Sanctuary AI's next-generation Phoenix robot, and more.

• Smart Reporting raised 23 million EUR to expand its medical reporting product: Smart Reporting raised €23 million to fund its global expansion, product development, and management team growth. Founded in 2014 as an LMU spin-off, their software enables automated, data-driven documentation for doctors. The company plans to grow by negotiating international partnerships.
• Chemix, an innovative Gen AI-powered battery developer, closed its $20M Series A: Chemix, an industry leader in battery development, leverages its proprietary MIX R&D platform to automate large swaths of its processes. The $20M Series A round led by Ibex Investors, allows Chemix to scale operations to meet customer demand and venture into more specialized segments.
• "AI-conic" is Kaffa Roastery's AI-developed coffee blend: Kaffa Roastery, a Helsinki coffee roaster, launched "AI-conic," a coffee blend developed in collaboration with AI consulting firm Elev. Elev provided LLMs with data from Kaffa's best-selling coffees to craft "a new exciting blend." The unconventional result was deemed perfect by blind tasters.
• Salesforge raised $500K in pre-seed capital to build a B2B sales AI copilot: Salesforge secured the pre-seed $500K to transform the B2B landscape with its AI-powered sales copilot. Salesforge creates email campaigns focused on personalization and deliverability, while Mailforge creates scalable email infrastructure in minutes.
• Parloa raised $66M for product development and market expansion: Parloa recently announced the successful closure of a $66 million Series B that will unlock product development innovations and drive market expansion. Parloa's AI platform empowers organizations to deliver personalized, dynamic, and human-like customer experiences while reducing wait times.
• Research from Adobe reveals most Americans believe generative AI will help them become more creative: Adobe's research reveals how people expect AI to help unlock their creativity and simplify their lives. Alongside online shopping, customer service, and travel, popular expected uses include research, writing, visuals, summarization, and coding.
• Treefera secured $12M to bring transparency and accountability to nature data: Treefera is an AI-enabled data platform bringing transparency and accuracy to nature-based asset reporting; providing, for instance, an auditable chain of custody for biochar production. The platform offers continuous monitoring, evaluates risks, and ensures process integrity.
• Augment Inc. came out of stealth announcing its $227M Series B: Augment raised a $227M Series B round. Its platform empowers developers with AI that understands codebases, supports teams, and protects IP. Augment was built with the belief that the best software development happens when AI empowers, rather than replaces, humans.
• Sanctuary AI introduced the seventh generation of its Phoenix humanoid robot: Vancouver-based Sanctuary AI unveiled the next generation of its general-purpose robot, Phoenix. Sanctuary AI is tackling the artificial general intelligence problem by mimicking human subsystems, and 7th-gen Phoenix is its most sophisticated human behavioral data capture technology to date.