Artificial Intelligence may spell the end for the SBIR/STTR Open Topic program if the Department of Defense doesn't intervene promptly.
In the realm of technology and innovation, the Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) programmes have long been a beacon for groundbreaking ideas. One such individual leading the charge in this domain is Jerry Ramey, managing principal for product and engineering at Dark Corner Solutions, who has also worked with the Air Force through the SBIR/STTR programme.
The author, an engineer and former chief technology officer, heads a small business that leverages machine learning and AI to accelerate the adoption of innovative ideas. However, the author notes a challenge in the current system: the high volume of nearly indistinguishable submissions, which often results in the best ideas not being chosen.
This issue extends beyond the SBIR/STTR programme, with the Defense Department facing a similar predicament in limiting the use of AI for proposals in its Open Topic programme. The National Institutes of Health has already taken action, banning the use of AI-generated content in proposals due to the overwhelming number of submissions.
The U.S. Air Force introduced Open Topic solicitations for its SBIR/STTR programme in 2018, changing the paradigm for acquiring new technologies from small businesses by allowing the government to occasionally tap into real-time industry ideas. However, the growing number of submissions has led to a situation where government SBIR/STTR proposal evaluators are drowning in a sea of cliches and em dashes.
To address this, the author suggests a human-centric approach. The DoD should demand submissions and reviews to be personal, with the use of videos, demos, exercises, and in-person meetings. This is not just about managing workload, but preserving the humanity in the process, as people are the innovators, aided by tools like AI.
The Air Force is suggested to lead the charge in restricting AI-generated content in SBIR/STTR proposals. The U.S. Department of Defense has already implemented guidelines restricting the use of artificial intelligence in SBIR/STTR proposals by setting specific ethical and security criteria, limiting certain AI applications, and requiring detailed documentation to ensure compliance with defense standards.
Despite the challenges, the number of total SBIR/STTR proposals increased every year from 2018 to 2022, with a brief dip in 2022. The introduction of ChatGPT at the end of 2022 has removed a previous barrier, allowing small businesses to submit more proposals. The author believes there will be a huge rally in the number of proposals when the numbers come out for the last few years.
However, it's worth noting that SBIR.gov has not published public data for SBIR/STTR proposals for the years following 2022, making it difficult to track the exact impact of these changes. By 2022, 40% of all awards for the SBIR/STTR programme were going to open-topic solicitations.
In conclusion, the integration of AI in the SBIR/STTR proposal process presents both opportunities and challenges. While it has made it easier for small businesses to submit proposals, it has also led to an overwhelming number of submissions, making the review process difficult. A human-centric approach, with a focus on personal interactions and experiences, could potentially address these issues and ensure that the best ideas rise to the top.
