Hedgeye Podcasts

Hedgeye NexGen: Building Wealth for Young & New Investors | Episode 8 | How AI Really Works

Feb 1, 2025
Ishmael Asad collaborates with Ryan Ricci, a Hedgeye Macro analyst, to unravel the complexities of artificial intelligence in finance. They discuss AI's historical roots, shedding light on neural networks and trial-and-error learning with coders and GPUs. The pair also contrasts generative and targeted AI in financial analyses, raising alarms about manipulation risks and biases in AI systems. The conversation culminates in the ethical dilemmas posed by AI in moral decision-making, stressing the need for human intuition in a technology-driven world.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

AI's Evolution

  • AI development has focused on increasing processing power rather than new code or ideas.
  • The core AI concepts from the 1980s remain largely unchanged, with advancements primarily driven by exponential growth in processing speed.
INSIGHT

AI as Neural Networks

  • AI models operate as neural networks with hidden layers processing inputs to generate outputs.
  • Coders define the inputs and evaluate the outputs, but the internal workings of the AI remain a black box.
INSIGHT

AI Learning

  • One way AI learns is through trial and error, similar to humans learning to catch.
  • Increased processing power allows AI to undergo vast amounts of trial and error rapidly, accelerating the learning process.
Get the Snipd Podcast app to discover more snips from this episode
Get the app