The stomping on the side of the kitchen table means that it requires a high degree of degree.
The data annotation industry pays hundreds of thousands of part-time contractors around the world to filter, rank and train AI's response to the world's largest AI companies. According to one technology CEO, who is doing the contract work.
Garrett Lord, CEO of Job Hunting and Handshake for AI Training Platform, said the data annotation industry is moving from the demands of generalists to the need for highly specialized mathematics and science experts.
“Currently, these models suck up the entire corpus of the internet and all the entire book and video,” he said in an episode of the “Grit” podcast released Monday. “They're getting better enough where they no longer need it like generalists.”
The Lord said that Frontier AI Labs need experts in the fields of accounting, law, medicine, and STEM domains in physics, mathematics, chemistry and more.
The CEO said contractors are making an average of over $100 or more than $125 an hour on the platform, and are applying domain expertise to AI training projects. Generalist contractors on other platforms range from a few dollars to about $40 per hour based on their challenges and locations.
The Lord's remarks come after a big shakeup with one of his handshake competitors. Scale AI recently received a $14.3 billion investment from Meta.
Just hours after Meta announced Big hit dealGoogle has suspended multiple projects with the company, BI reported last month. Openai and Elon Musk's Xai have also suspended projects of several sizes, and contractors of the size working on them told BI.
Other data labeling platforms such as Handshake and Appen, Prolific and Turing are welcoming this deal. Executives from these companies said they are attracting more attention from clients in major technology.
“The lab doesn't want to know what data other labs are using to improve their models,” Lord said in an interview with Time Magazine last month. He added that demand for handshake services “trimmed overnight” as a result of meta transactions.
“For General Motors or Toyota, I don't want to see how competitors enter the manufacturing plant and carry out the process,” he told time.
A Scale spokesman told BI in a statement last month. Customer Data Protection.
“Security and customer trust are always at the heart of our business and we continue to ensure that there is adequate protection in place to protect all our work with our customers,” the statement said.
The handshake did not immediately respond to requests for comment.

