Louisiana Startup Cuts AI Costs—No Extra Hardware Needed
Training an artificial intelligence is notoriously expensive, not just in terms of the vast electricity use, but also in the top-of-the-line hardware it requires. However, a New Orleans-based AI company is poised to revolutionize the costs. NOLA AI has developed a new algorithm called AToMIC that it says can train AI more efficiently, with fewer "hallucinations" and lower costs, all without needing a single extra CPU.
Across the country, power-hungry AI data facilities are being built with tons of expensive hardware. NOLA AI has created a new AI framework that they say creates a more "human-like" AI model by introducing the ability to understand context better, read emotion, and understand nuance.
The Automated Theory of Mind in Communication framework won't just make AI smarter; it will reduce the costs and electricity consumption AI training requires. According to NOLA AI, the AToMIC framework will cut back on hardware costs through fine-tuned measurements during the training process, giving the software a better idea of how much power is being consumed, how accurate the AI can be, and the rationality of it as well.
With Artificial Intelligence becoming a booming industry in the United States, this is a huge win for Louisiana, a state that has been economically a bit behind others. Meta has shown interest in creating an AI data center in Louisiana, and with an AI development firm located right here in the state, it gives the Bayou State a foot in the door of the fastest-growing technology industry.