Information technology engineers and chief information officers for major companies around the country are struggling to keep up with the demand for artificial intelligence (AI) applications.
- The explosive rollout of OpenAI’s ChatGPT on November 30 has changed the world and begun what is being called the AI arms race to create the best and most versatile AI.
- The rush is creating a high demand for new AI applications, and software developers are struggling to meet that demand.
- Amerisave Mortgage Corp. chief information officer Magesh Sarma, speaking with The Wall Street Journal, says the time and cost of building new AI models is too intensive to meet the market demand.
- Many requests for cost-saving AI are simply being rejected due to the scope and intensiveness of many of his projects. The industry is struggling to tackle all of the issues being pitched to them.
- However, the demand has been a benefit to his team. In a time of frequent tech mass-layoffs, keeping his 76-member team intact.
Why It’s Important
AIs stand to become more prominent in the coming months and years, as tech firms like Microsoft, Google, and Baidu rush to front chatbot features in their search engines and suggestive text features. The technology is still unpolished, as evidenced by early failures with data retrieval and factual accuracy, but the demand is going to progressively automate more processes to the machines.
The success of the technology only stands to embolden companies to further lean into AI technologies, building more smaller programs that can automate smaller tasks and make companies more efficient. This is bad news for engineers like Sarma, who need time to build these programs out and are suddenly swamped with requests for smaller projects.
Backing Up A Bit
One of Amerisave’s most prominent recent projects was a multi-year project designed to assist with underwriting loans that cost between $20 million and $30 million, primarily in labor costs. The process is intensive, requiring multiple algorithms that handle different tasks, and it takes time to build and adjust them.
The expense is added by the challenge that very few engineers are qualified to handle this work, meaning the median salary for data scientists exceeds $156,000, according to The Wall Street Journal.
The effort paid off, giving the company the ability to expand from less than $2 billion in mortgage loans to more than $24 billion in two years.