Open Source AI is Leading to Breakthroughs in Healthcare, Education and Entrepreneurship

Takeaways

• Open source AI is transforming industries by providing accessible and transparent tools that allow for collaboration and ingenuity.
• Researchers and developers across the world are using Llama to solve real-world issues, from gaps in clinical cancer trials to inefficiencies in agriculture.

We believe open source AI is the path forward, and we’re thrilled that developers around the world are already using open source models like Llama to shape the future.

This year, we’ve held a series of events to build an open source community and support Llama developers through mentorship, funding and a broad network. These developers have launched innovative projects that span science, education, healthcare, entrepreneurship and more. Their projects exhibit the unique advantages of open source AI and demonstrate how transparent and accessible source code allows for collaboration, ingenuity and flexibility – the exact qualities needed for safe and stable development in AI. 

Here’s a selection of stories about how these developers are using open source AI to solve pressing global issues.

City of Hope

At the City of Hope National Medical Center, researchers are using LLMs to build tools for cancer doctors, including a tool that matches cancer patients with relevant clinical trials.

“Medicine is one of the best, most urgent applications for LLMs, given the scale and sensitivity of the text data generated in the clinic.”

– Kingson Man, Principal Data Scientist at City of Hope

Man explains that 40% of cancer clinical trials fail because they can’t find enough eligible patients to participate, and only 3-7% of patients make it into a trial. To address this gap, City of Hope built a matching tool that analyzes whether specific clinical trial eligibility criteria apply to current City of Hope patients, then produces a matching score. With that matching score, doctors can surface additional, possibly overlooked clinical trials for consideration for their patients.

Gene Outlook

OtonoCo uses open source LLMs to power Gene Outlook, a platform designed to accelerate discovery and extraction of insights from data, particularly in the field of genomics. 

Researchers use Gene Outlook to find important gene signatures or other kinds of gene-related data in cancer patients, then Llama turns the gene symbol into content that scientists and businesses can easily understand. OtonCo has also developed an organism-aware LLM on top of Llama called GeneTuned LLM, and is in the process of expanding Gene Outlook to the areas of gut health and microbiome.

“With open source, we don’t have to worry about creating a new AI algorithm. We can just focus on the problem and work backwards to see what open source technology is suitable to solve that problem.”

– Jong Hang Siong, Founder and Chairman of OtonoCo

DLYog Lab

DLYog Lab was co-founded by Tarun Chawdhury and Mousumi Chawdhury, who were inspired by their experience with their son, who has special needs. Together they developed a GenAI-powered app that simplifies the process of creating Individualized Education Plans (IEPs) for special education. 

The app processes transcripts from IEP meetings with parents, their child, teachers and specialists to generate an education plan detailing goals and services for the child.

“Llama helps small players like us realize our dream for social good. It was created for my son’s benefit but now we can extend this to many other parents.”

– Tarun Chawdhury, Co-founder of DLYog Lab

DLYog’s app enables teachers to create personalized plans quickly based on a child’s profile, and Llama’s publicly accessible open source architecture can be rapidly scaled for small institutions like private schools. Because the app can be deployed in a private data center with minimal infrastructure requirements, small schools and private institutions with concerns about data security can adopt it within their own perimeters. And with the newest version of Llama, DLYog Lab’s app now analyzes nonverbal cues like facial expressions and eye movements, enabling insights that go beyond traditional approaches to IEPs.   

Pratham Education Foundation

Pratham Education Foundation (PEF), located in India, used Llama to build a solution for young mothers to learn about childcare and early childhood education. This information can be difficult to access on the open web, which often provides confusing or conflicting information.

“Social good must be open source. As a non-revenue-generating organization, it’s important for us to get the most value for our dollars. That’s another way Llama has been really helpful.”

– Nishant Baghel, Director of Technology Innovation at PEF

PEF’s WhatsApp-based chatbot spans 12 Indian languages, and mothers can interact with the bot by voice, rather than typing out text. The bot generates an audio and video answer based on verified documents, in addition sending a video link that makes it easier for mothers to digest needed information. More than 40,000 mothers in India have used the product. 

According to Nishant Baghel, PEF’s Director of Technology Innovation, Llama acts as PEF’s “backbone” to generate and consume knowledge.

HiiiWAV 

HiiiWAV is an incubator based in Oakland, California, teaching artists, particularly Black musicians, how to build startups using AI. Their goal is to help these artists see – and harness – the power of AI and machine learning.

Because it’s important for early-stage startups to keep costs low, HiiiWAV recommends open source technology, which levels the playing field by expanding access to often expensive, highly technical tools. Among available options, they call Llama “top of the list,” because it not only cuts costs but also enables them to build community and share information freely.

“Off-the-shelf closed-source options get pricey with usage, and they’re hard to build with. We’re also building AI-powered hardware, so we need the control and freedom to customize that’s missing from closed source options. Open source is definitely the way to go.”

– Bosko Kante, Executive Director of HiiiWAV

Entrepreneurs in the HiiiWAV incubator have used open source AI to build a variety of tools, including GooRoo, a mobile voice-assisted record producer and Choice Scores, an AI-powered platform that turns hip-hop songs into musical scores.

AIIRA

AIIRA, the AI Institute for Resilient Agriculture, builds AI tools to help agriculturists with their daily operations. Their aim is to encourage AI adoption among farmers to help them solve challenges in crop improvement and production. Llama’s open source technology allows for multilingual, geographic and context-dependent fine-tuning so that the organization can have different models that work for particular regions. AIIRA uses Llama to generate personalized recommendations to farmers to address pests and weeds.

“Open source AI’s ability to deploy at scale is transformative. It works in the United States, it can work in Africa, in India, across the globe. I’m hoping that this sort of open source spirit continues in the future so that we can continue to leverage the latest advances in AI.”

– Chinmay Hegde, AIIRA member and professor at New York University

These innovations with Llama show how open source AI is positively impacting the world, and we look forward to seeing the global community continue to realize the potential of open source AI.