For a small startup or nonprofit, building a technology platform can be a daunting experience. You have limited funds and are faced with challenges and dilemmas–what you build; what tech stack you use to build it; what infrastructure you build it on; hiring or out-sourcing, near-shore or off-shore, remote or local. These choices can be overwhelming. I have seen first-hand a poor design decision made at inception preventing an organization from meeting or exceeding the aspirations of the founders.
Building the technology platform is the single biggest investment a startup will make, and you only get one opportunity to make it.
In the case of a company where technology is the product, building the technology platform is the single biggest investment a startup will make, and you only get one opportunity to make it. If the technology platform fails, the company fails.
This is why seasoned leadership is critical to navigate decisions that will make or break a company. But deep experience is not always affordable. Imagine having to decide between an experienced CTO or three developers. For a startup that is a no-brainer—pick the developers to get you to your minimal viable product (MVP). But the implications of not engaging a CTO at the right time to set the course and strategize for the future could be catastrophic.
From the onset, a startup should have the absolute best it can afford in the following areas:
- Robust technology foundation
- First-class technology leadership
- Leading-edge innovation
- Exquisite design and experience
Our commitment to our Stead Impact community is to provide this leadership as part of our suite of services. There is nothing more fulfilling for me than to bridge temporary gaps in leadership for the founders and leaders within our portfolio to support your success.
AI and Impact
Artificial Intelligence is a term that generates both hype and skepticism; humans like to be in control and AI removes that control. Whether it’s the apocalyptic scenes of Terminator, inaccuracies through hallucinations, or concern about job displacement, these concerns dominate the discussion in both small and large businesses.
There are many branches of AI that provide different value propositions. Natural language interpretation, translation, facial recognition, object recognition, all are employed in our mobile devices–a quick search for “dog” in your photo library confirms this if you don’t believe me. So, AI is mainstream in this sense, but what is happening on the business-to-business level?
Over many years we have seen technologies deployed to solve healthcare problems. But Generative AI has placed us at a technological inflection point that promises the single biggest change since the world-wide-web in the 90s. We are going to see some big problems solved through technology for the benefit of millions of people around the world.
The Stead Impact Ventures portfolio is using AI for positive healthcare impact. For example, generative AI can be trained on high-quality reference sources, ensuring accuracy. A surgeon could ask a series of questions about a difficult case and generative AI could deliver responses from millions of pieces of vetted literature from multiple medical specialties. Dr. House would have loved a tool like that, but the TV show would not have been as interesting. Imagine the improved patient outcome from a faster and more accurate diagnosis and treatment selection. That is the power and potential of AI. And we are already doing this with Dr. Aaron Cohen-Gadol at Atlas Meditech with AtlasGPT.
Another example is generative AI used to educate someone on Alzheimer’s disease after a parent is diagnosed. The AI responses can be calibrated to the user’s learning level, so as they understand more, the level of information becomes more complex. Not only that, but the responses can be delivered in a compassionate tone. This is something that can’t be achieved through reading material posted on the general web. This is exactly what TapRoot is developing with its elbi app to help family members who are caring for loved ones with dementia.
As for job displacement, some jobs will go away, but just as in the technological revolution in the 2000s, computers simply generated more jobs. The same will likely happen with AI, which will allow us to relegate repetitive tasks and focus on the creative and critical thinking that we do best.
Can We Trust AI?
Shortly after Chat GPT was launched, there was a backlash about some of its shortcomings. We became familiar with the term hallucinations. It is true that this can be a significant drawback. Imagine training a large language model, the model that underpins generative AI solutions, with un-validated information from the internet. Would you trust this? Well, this is exactly what Chat GPT is using as its source of information. So, it is little wonder that it has limitations.
By training AI using verified data, such as peer-reviewed content, we can remove hallucinations. This is exactly the approach used in Atlas GPT and elbi, providing accurate results and confidence in the user.
AI Regulation
These technologies are outpacing the regulators and legal system. With AI providing the ability to make a medical diagnosis, who is responsible for a misdiagnosis? The regulators, not only in the U.S., but in other parts of the world, are wrestling with this challenge. The EU launched the AI Act, and the U.S. was not far behind with the FDA Artificial Intelligence and Machine Learning (AI/ML) for Drug Development, but California’s recent veto clearly shows that views are divided. These entities are waking up to the need to apply some level of restraint on AI. My hope is these regulations lay the groundwork to prevent misuse but not impinge potential healthcare breakthroughs.
Optimism for AI
We are working with our Stead Impact community to use AI in our mission to improve health and wellbeing for people globally. Each day I’m inspired by stories large and small of how people are using AI to solve problems. A colleague of mine shared a story of college students who recently completed a pilot program at a Kenyan refugee camp. Using AI, the one teacher assigned to 100 multiage students was able to develop individually calibrated curriculum in English and Swahili. I truly believe technology and AI are a tool for us to find new ways to solve problems, help others, and do good in the world.
