Welcome to the AI Ecosystem Report, featuring practitioner analyst and entrepreneur Toni Witt. This series is intended to deliver the timely intelligence about artificial intelligence (AI) you need to get up to speed for an upcoming client engagement or board meeting.
Highlights
Innovation (00:50)
Fast Company covered AI’s role in the real estate industry. For context, a real estate analyst is responsible for analyzing real estate deals, forecasting profits and economics before deals happen. With many of their tasks being repetitive, there is an opportunity for AI to add value to this role.
Real estate investment software company TermSheets released ETHAN, an AI-based real estate analyst. It can quickly create investment reports, aggregate datasets, and answer specific questions. Similarly, there has been a growing number of innovations, noted in the Fast Company article, being built out internally by large firms in this industry:
- JLL, one of the largest real estate firms in the world, built JLL GPT, an AI assistant trained on its enormous dataset of documents, including leases, facilities management policies, contracts, and more. JLL GPT helps with communicating with tenants, identifying good investment opportunities, and streamlining work order filing processes.
- Skanska, one of the largest construction and engineering companies in the world, built its own AI chatbot for internal use. The chatbot pulls data from across a construction project to help managers make decisions and stay on top of timelines.
- Slate is a startup that built an AI tool that pulls data from various sources to predict construction errors or delays that might occur. By doing so, it’s able to proactively warn users against them and make suggestions to hedge the negative effects and stay on the timeline.
The AI Ecosystem Q1 2024 Report compiles the innovations, funding, and products highlighted in AI Ecosystem Reports from the first quarter of 2024. Download now for perspectives on the companies, investments, innovations, and solutions shaping the future of AI.
Funding (04:50)
At Sequoia Captial’s AI Summit, the company pointed out that it has invested nearly $50 billion into the chips needed to run large language models (LLMs). However, revenues from GenAI startups have only hit around $3 billion. While ChatGPT was a huge hit and OpenAI is generating serious revenue, other LLM and GenAI startups haven’t reached high revenue levels.
New businesses have to solve a need better than their competition. They have to apply tech in a way that adds to the bottom line. The big picture trend unfolding in the startup funding world is that funding is moving slowly away from GenAI platforms and LLMs and moving toward vertical-specific applications of existing models. This is increasingly where venture and corporate venture capital is flowing to.
Solution of the Week (06:47)
There has been a rise in small language models (SLMs), which are cheaper to build, train, and run. Microsoft announced a new model: Phi-3-mini. This is the newest and smallest in a series of three models that the company released. The Phi-3-mini has only 3.8 billion parameters.
Despite having very few parameters compared to other models, after undergoing performance tests, Phi-3-mini has been found to perform nearly as well as GPT 3.5. The Microsoft team was able to achieve this partially due to training it on curated, high-quality data. This lightweight, energy-efficient model can run on consumer devices, including mobile phones and laptops, without an internet connection.
Phi-3-mini is available now through Azure and Hugging Face.
Ask AI Ecosystem Copilot about this analysis