Algorithms are progressing so fast, an AI researcher got spooked.
Original headline: Anthropic can already retrain Mythos with "a fraction of the compute originally needed," says former researcher Coxon — the-decoder.com
More in AI and machine learning news
A former OpenAI and Anthropic researcher, Jacob Coxon, left his role at Anthropic after seeing a chart from the company's IPO roadshow. The chart illustrated algorithmic efficiency improvements in AI training, showing compute requirements dropping at an accelerating rate. Coxon was alarmed by how steep the curve was, noting that advancements aren't just fast, but speeding up.
He points to Anthropic's 'Mythos' model as an example, stating the company could now retrain it with only a fraction of the original computational cost. His key takeaway is that this runaway progress could consolidate power among a few leaders, prompting his call for a development pause to let competitors catch up and 'commoditize human-level intelligence.'
Read the original at the-decoder.com
Satirical commentary on real tech news, summarised with AI. Read the original source for the facts. How ZeroDaily works