Small Businesses Face Uphill Battle in AI Race
Small and medium-sized businesses will struggle to keep pace with tech giants like OpenAI in developing their own artificial intelligence (AI) models, according to a new report from Stanford University.
In an interview, Nestor Maslej, the editor-in-chief of Stanford’s newly released 2024 AI Index Report, highlighted the study’s findings on the growing AI divide between large and small companies. While tech behemoths pour billions into AI R&D, smaller firms lack the resources and talent to compete head-on.
“A small or even medium-sized business will not be able to train a frontier foundation model that can compete with the likes of GPT-4, Gemini or Claude,” Maslej said. “However, there are some fairly competent open-source models, such as Llama 2 and Mistral, that are freely accessible. A lot can be done with these kinds of open-source models, and they are likely to continue improving over time. In a few years, there may be an open, relatively low-parameter model that works as well as GPT-4 does today.”
A study from PYMNTS last year highlighted that generative AI technologies such as OpenAI’s ChatGPT could significantly enhance productivity, yet they also risk disrupting employment patterns.
Surging Costs of AI
A major takeaway from the report is the possible disconnect between AI benchmarks and actual business requirements in the real world.
“To me, it is less about improving the models on these tasks and more about asking whether the benchmarks we have are even well-suited to evaluate the business utility of these systems,” Maslej stated. “The current benchmarks may not be well-aligned with the real-world needs of businesses.”
The report indicated that while private investment in AI generally declined last year, funding for generative AI experienced a dramatic surge, growing nearly eightfold from 2022 to $25.2 billion. Leading players in the generative AI industry, including OpenAI, Anthropic, Hugging Face and Inflection, reported substantial increases in their fundraising efforts.
Maslej highlighted that while the costs of adopting AI are considerable, they are overshadowed by the expenses associated with training the systems.
“Adoption is less of a cost problem because the real cost lies in training the systems. Most companies do not need to worry about training their own models and can instead adopt existing models, which are available either freely through open source or through relatively cost-accessible APIs,” he explained.
Pushing Standards for AI
The report also calls for standardized benchmarks in responsible AI development. Maslej imagines a future where common benchmarks allow businesses to easily compare and choose AI models that match their ethical standards. “Standardization would make it simpler for businesses to more confidently ascertain how various AI models compare to one another,” he stated.
Balancing profit with ethical concerns emerges as a key challenge. The report shows that while many businesses are concerned about issues like privacy and data governance, fewer are taking concrete steps to mitigate these risks. “The more pressing question is whether businesses are actually taking steps to address some of these concerns,” Maslej noted.
Measuring AI’s impact on worker productivity across different industries remains complex. “It is possible to measure productivity within various industries; however, comparing productivity gains across industries is more challenging,” Maslej said.
Looking ahead, the report highlights the need for businesses to navigate an increasingly complex regulatory landscape. On Tuesday, Utah Sen. Mitt Romney and several Senate colleagues unveiled a plan to guard against the potential dangers of AI. These include threats in biological, chemical, cyber and nuclear areas by increasing federal regulation of advanced technological developments.
Maslej emphasized the importance of staying vigilant. “Navigating this issue will be challenging. The regulatory standards for AI are still unclear.”
As public awareness of AI grows, Maslej believes that businesses must address concerns about job displacement and data privacy. “As people become more aware of AI, how can businesses proactively address nervousness, especially regarding job displacement and data privacy?” he posed as a crucial question for the industry to consider.
The 2024 AI Index Report is meant to guide businesses and society in navigating the rapid advancements in artificial intelligence. Maslej concluded, “The AI landscape is evolving at an unprecedented pace, presenting both immense opportunities and daunting challenges.”
