Photo: Microscope close-up / iStock, Kolosov
On May 1, IBM shocked the tech industry by suspending the hiring of 7,800 positions it determined could be better served by artificial intelligence.
Many biopharmaceutical companies are rapidly adopting artificial intelligence (AI) and machine learning (ML) for many applications, from preventive medicine to drug discovery. But it remains to be seen whether the biotech industry will suffer the same purge as the high-tech industry, said Jeff Morton, a biotech patent attorney and partner at Procopio Corey Hargreaves & Savitch.
“The jury is still out on how AI will affect things,” he said. bio space. Morton said data analysis could be increasingly automated, but ultimately companies still need to prove that drugs and treatments work for patients. This will require complex and creative research, and “I think it will still rely on humans,” he added.
“Most of the people I talk to on a daily basis in the biotech space are senior scientists, who are likely to enjoy greater job security as AI becomes more integrated into biopharmaceuticals,” Morton said. They are high officials,” he said. But people in other positions, or students considering a career in biotechnology, might take a different view, he said.
Morton said the role of AI in biotechnology is wide-ranging and heterogeneous, meaning that different sectors may be affected differently. AI algorithms have the potential to discover new targets for existing drugs, accelerate drug development, screen clinical and research data, monitor patients, and analyze market trends.
AI in biotechnology is nothing new, but generative AI has received media attention in recent months, the most notable example being GPT-4, the engine that powers ChatGPT. Generative AI can use existing data to generate original content. GPT-4 has demonstrated its ability to perform a variety of tasks and generate new content such as working code and legal summaries.
human atmosphere
eClinical Solutions CEO Raj Indupuri said: bio space He said he expects generative AI to change many roles in biotechnology, but he is skeptical that AI will replace humans in clinical research and drug development. Companies, especially in life sciences, need human actors to validate the output of algorithms, he said.
“We still need to validate what AI and ML are predicting,” says Indupuri. “The consequences of getting something wrong are serious.”
In other words, AI makes mistakes, contains biases, and requires validation and thoughtful implementation. Generative AI is particularly difficult to validate, so human input will be required at multiple stages of the clinical development process in the foreseeable future, Indapuri said.
AI algorithms learn from existing data and classify or generate new data based on known data. But “clinical development is so complex that they have little in common,” Indupuri said. This means that it is difficult for algorithms to generalize data from past successful trials and produce accurate output, especially for rare diseases with small datasets.
Therefore, companies still need research teams to analyze data and make decisions at each stage of the drug development process, from early drug research to clinical studies, Indapuri said.
He added that AI can eliminate “inefficiencies” in drug discovery and data analysis. This could result in job losses, but likely “not as much” as in technology, he said.
Light of hope
Indapuri added that AI could also create many new roles or simply change old roles. In addition, many manual tasks “can be eliminated or transformed using AI,” he said, so some may report higher levels of job satisfaction.
As the use of artificial intelligence increases, meeting future labor demands will require upskilling—training existing employees to work with AI and ML technologies, Indapuri said. Mr. Morton agreed.
“There’s already been a lot of retraining,” Morton said.
Still, both he and Indapuri said they believe AI will play a major role in the biopharmaceutical space in the coming years.
“There seems to be a general consensus that things will change, but it’s hard to predict how that will change,” Morton said.
