AI @ UofL
Artificial Intelligence (AI) is a transformative technology field, encompassing machine learning (ML) and generative AI (GenAI), with the potential to revolutionize numerous aspects of our lives.
Generative artificial intelligence (GenAI) at UofL
Introduction for the UofL community on the best practice for generative artificial intelligence (GenAI) as part of teaching, learning and administrative use.
GenAI options
Microsoft Copilot (UofL Approved)
An everyday AI companion, this free tool is part of UofL’s Office365 package and integrated into each Office software, Windows operating systems and their web tool, Edge. Copilot gives you the advanced power of OpenAI’s top models without a subscription, making it the most accessible and value-packed for high-quality answers and content creation.
Microsoft 365 Copilot (UofL Approved)
The unrivaled assistant for professionals, this paid resource offers unmatched integration into your daily workflow. By securely connecting with your emails, meetings and documents within the Microsoft 365 apps you use every day, it transforms from a simple chatbot into a personalized work partner that actively manages your tasks and productivity. To add-on M365 Copilot to your UofL free subscription, employees should contact our HelpDesk.
Gemini Advanced (Google)
Google’s most capable model (1.5 Pro or version 3) is a strong contender for users deeply embedded in the Google ecosystem. It excels at complex multimodal tasks, such as analyzing video and text simultaneously, and offers robust reasoning capabilities for coding and creative projects.
Claude 3.5 Sonnet (Anthropic)
A favorite among developers, this model is widely regarded for its exceptional coding abilities and nuanced, human-like writing. It features a massive context window for analyzing long documents and avoids the robotic tone often found in other AI tools.
Perplexity
A specialized "conversational answer engine" designed for accuracy rather than creative chat. It uses Generative AI to perform real-time web searches and provides direct footnotes for every claim, making it a superior tool specifically for fact-checking and research.
ChatGPT (OpenAI)
The popular chatbot that started the wave, powered by GPT-4o. It remains a versatile all rounder for brainstorming, voice interaction, and drafting, though it lacks the deep operating system integration found in Microsoft's tools
Imagen 3 (Google)
Available via Gemini and ImageFX, this is Google’s highest-quality model for generating photorealistic visuals. It offers exceptional prompt adherence and handles text rendering within images better than previous iterations.
Flux.2 (Black Forest Labs)
A community favorite known for producing stunningly realistic images with complex lighting and accurate human anatomy. It is often used by creators looking for a balance between high fidelity and artistic control.
Midjourney v6
The industry standard for artistic and stylized image generation, producing high-fidelity, painterly and cinematic visuals. It operates primarily through Discord or a web alpha and is the go-to for creative professionals.
Ideogram
A specialized generator that solves the "text problem" in AI art by accurately rendering typography and lettering. It is the ideal tool for creating logos, t-shirt designs and posters where legibility is required.
Recraft
A unique design tool that generates vector graphics (SVG) rather than just static pixels. It is perfect for professional designers needing scalable icons, logos and illustrations that fit specific brand style guidelines.
DALL-E 3 (OpenAI)
Integrated into ChatGPT and Microsoft tools, this model is beginner-friendly and excellent for following complex instructions. However, it offers less granular control over specific artistic styles compared to Flux or Midjourney.
Google Veo (Flow)
Google’s generative video model, available in select workspace labs, is designed to create 1080p+ cinematic clips. It interprets prompts to produce consistent, high-quality motion suitable for creative projects.
Genmo
Genmo is an AI platform using open-source models that generates high-quality videos, images and 3D models with customization prompts and animation tools.
Kling AI
A newer model is gaining popularity for its ability to generate longer clips (5-10 seconds) with high motion coherence. It produces realistic physical movements, making it a strong competitor to established video platforms
Luma Dream Machine
A fast and accessible platform that turns text or images into high-quality 5-second video clips in minutes. It is excellent for quick prototyping, meme creation and animating static ideas without complex software.
Sora (OpenAI)
OpenAI’s advanced text-to-video model is capable of generating complex clips with multiple subjects and specific actions. While highly impressive, it currently has limited public availability compared to other tools on this list.
NotebookLM (Google)
A revolutionary "AI research notebook" that learns from your uploaded documents to become an expert on your specific sources. It generates summaries, citations and even audio discussions, ensuring answers are grounded in your material.
SciSpace
An AI copilot designed to decode scientific papers and explain complex math or text in simple language. It helps researchers quickly find relevant literature and understand key concepts with citation-backed explanations.
Consensus
A search engine for science that extracts findings from peer-reviewed papers to answer questions with evidence. It provides summaries based on data rather than hallucinations, making it essential for academic verification.
Elicit
Automates the literature review process by finding relevant papers and extracting key data like methodology and findings. It organizes this information into comparison tables, saving researchers hours of manual sorting.
ResearchRabbit
ResearchRabbit is a citation-mapping tool that helps researchers explore, organize and connect academic papers and authors. It works through literature citation networks based on relevant research interests.