Today's Edition

• Study shows that the way the brain learns is different from the way that artificial intelligence systems learn

• AI Stocks: Best Artificial Intelligence Stocks To Watch Amid ChatGPT Hype

• Quote of the Day

• Tech Spotlight

• On This Day In History

📰 Study shows that the way the brain learns is different from the way that artificial intelligence systems learn

Researchers from the MRC Brain Network Dynamics Unit and Oxford University’s Department of Computer Science have introduced a new principle to understand how the brain modifies neuronal connections during learning. This insight could impact future brain network studies and enhance learning algorithms in artificial intelligence (AI).

Learning involves identifying which elements in the information process cause output errors. In AI, this is usually done through backpropagation, adjusting parameters to minimize output error. This principle is thought to be similar to brain learning.

However, the human brain outperforms current AI systems. For instance, humans can learn from a single exposure, whereas AI requires repeated training. Additionally, humans retain existing knowledge while learning new information, unlike AI systems where new learning can degrade previous knowledge.

This led researchers to explore the brain's fundamental learning principle. They examined mathematical models of neuronal behavior and synaptic connections, discovering a different learning principle from that in AI.

In AI, an external algorithm alters synaptic connections to reduce error. In contrast, the researchers suggest the human brain first optimizes neuronal activity before modifying synapses, a process termed 'prospective configuration'. This approach, detailed in Nature Neuroscience, preserves existing knowledge, facilitating faster learning.

The researchers illustrate this with a bear fishing analogy. In AI, if a bear can't hear a river due to a damaged ear, it might also 'forget' the smell of salmon, leading to a wrong conclusion of no salmon. In contrast, an animal brain would still recognize the salmon's smell despite the lack of sound, indicating salmon presence.

The team developed a mathematical theory showing that prospective configuration minimizes learning interference. This theory better explains neural activity and behavior in learning experiments than traditional AI networks.

Professor Rafal Bogacz, the lead researcher, notes a significant gap between prospective configuration models and detailed brain network anatomy. The team aims to bridge this gap, understanding how prospective configuration operates in cortical networks.

Dr. Yuhang Song, the study's first author, remarks that simulating prospective configuration on current computers is slow and inefficient due to fundamental differences from the biological brain. Developing new computers or dedicated hardware inspired by the brain could enable rapid, energy-efficient implementation of prospective configuration.

Read more here

📰 AI Stocks: Best Artificial Intelligence Stocks To Watch Amid ChatGPT Hype

As investor excitement around artificial intelligence intensifies, numerous businesses are now promoting AI-focused product strategies. However, it's not easy for investors to identify authentic AI stocks that are already generating revenue from generative AI. Examples of such companies include Microsoft (MSFT) and Nvidia (NVDA). For others, like Google's parent company Alphabet (GOOGL), the emergence of generative AI presents both potential risks and opportunities.

For those seeking to invest in the top AI stocks, exercising caution in this hype-driven environment is advisable. Generally, it's wise to target AI stocks that leverage artificial intelligence to enhance their products or to secure a competitive advantage.

Read more here

🗣️ AI “Quote of the Day”

"Every leap in knowledge is a whisper from the future, revealing a tapestry where threads of curiosity intertwine with the fabric of discovery."

Created with ChatGPT / DALL-E

🚀 AI Tech Spotlight

🔍️ RhetorAI - Automate customer interviews

RhetorAI is a feedback collection AI tool that uses ChatGPT to interview customers for businesses. It offers automated follow-ups, multilingual support, insights generation, customizable conversation lengths, website embedding, and email notifications. Users can test the service with demos for various scenarios. RhetorAI emphasizes privacy, not using customer data to train their AI. They offer various pricing plans, including a free basic plan, a plus plan for teams, and custom solutions for enterprises. The service is trusted by over 1300 users and aims to facilitate customer-driven decisions with stored transcripts and profound insights.

🗓️ On This Day In History

During this month in 1610, Italian astronomer Galileo made the earthshaking discoveries that four moons revolve around Jupiter and that the telescope reveals many more stars than are visible to the naked eye.