Ben Fielding: Neural architecture search automates deep learning, the shift to horizontal scaling is essential, and blockchain security enhances consensus algorithms | Unchained

Key Takeaways

• Neural architecture search automates the creation of deep neural networks, enhancing efficiency in machine learning.
• Current machine learning training methods lack scalability compared to evolutionary algorithms.
• Transitioning from vertical to horizontal scaling is crucial for improving machine learning efficiency.
• Machine learning is poised for a transformative shift similar to the MapReduce moment in computing.
• Existing blockchain systems offer security for new consensus algorithms without starting from scratch.
• Consensus algorithms can resolve disputes between devices autonomously, enhancing automation.
• Smart contracts facilitate immediate dispute resolution and verification in machine learning transactions.
• Wallet addresses in crypto serve as identities, which machine learning models can control for transactions.
• Trust is fundamental to the functioning of crypto systems, ensuring stability and reliability.
• Verification of machine learning model execution is essential for trust and dispute resolution in decentralized systems.

Guest intro

Ben Fielding is CEO and co-founder of Gensyn, the decentralized machine learning compute protocol. He holds a PhD in neural architecture search for deep learning and computer vision. Previously, he co-founded a data privacy startup.

The role of neural architecture search in AI

• Neural architecture search automates deep neural network creation, optimizing structure during training.

• I focused my entire research on that problem… it’s an area called neural architecture search

  — Ben Fielding

• Automating model creation enhances efficiency in machine learning.
• Current machine learning methods are limited in scalability compared to evolutionary algorithms.

• These techniques would scale in a way that the centralized techniques don’t scale

  — Ben Fielding

• Distributed approaches offer a path for improvement over traditional methods.
• Understanding the significance of automating deep learning model creation is crucial.
• This innovation can significantly enhance model development efficiency.

The shift from vertical to horizontal scaling in machine learning

• Machine learning needs to transition from vertical to horizontal scaling for greater efficiency.

• We as a company have a deep belief that machine learning needs a horizontal scalability moment

  — Ben Fielding

• Vertical scaling involves adding more compute power, while horizontal scaling distributes tasks.
• Horizontal scaling allows for continued scaling across multiple devices.
• This shift is analogous to Google’s MapReduce in computing.

• Our belief is that machine learning is ready for its mapreduce moment

  — Ben Fielding

• The MapReduce moment represents a significant shift in machine learning design.
• Implementing horizontal scaling can enhance machine learning infrastructure.

Leveraging blockchain security for consensus algorithms

• Existing blockchain systems provide necessary security for new consensus algorithms.

• This class of existing crypto networks provides that security

  — Ben Fielding

• New algorithms can be developed without bootstrapping security from scratch.
• Blockchain security is a critical mechanism for integrating AI and blockchain.
• Consensus algorithms can resolve disputes between devices without human intervention.

• The reason we even discovered this technology was from research papers

  — Ben Fielding

• Automating processes in AI and blockchain integration relies on consensus algorithms.
• Understanding blockchain security’s role is essential for developing new technologies.

Smart contracts in machine learning transactions

• Smart contracts enable instantaneous dispute resolution and verification.

• That’s what smart contracts give us… they give us the way to define a very specific kind of exchange

  — Ben Fielding

• They automate and secure transactions in machine learning.
• Enhancing efficiency and trust in machine learning operations is a critical function of smart contracts.
• Smart contracts execute verification and arbitration almost instantaneously.
• They play a vital role in automating machine learning transactions.
• Understanding smart contracts’ role is crucial for leveraging blockchain in AI.
• The use of smart contracts can streamline machine learning processes.

Machine learning models and crypto identity

• A wallet address in crypto serves as an identity, controlled by machine learning models.

• A wallet address is an identity within the crypto world

  — Ben Fielding

• Machine learning models can facilitate transactions by controlling these addresses.
• This intersection highlights a foundational concept in technology application.
• Trust is the absolute key for crypto systems’ functioning.

• Imagine if kind of crypto existed but it didn’t have trust… the whole thing would just fall apart

  — Ben Fielding

• Trust ensures the stability and reliability of decentralized systems.
• Verification of machine learning model execution is crucial for enabling trust.

Verification and trust in decentralized systems

• Verification of model execution enhances trust and dispute resolution.

• The ability to take a machine learning model execution and verify it at the consensus of the nodes

  — Ben Fielding

• Verification processes are essential for decentralized applications’ functionality.
• Trust is a critical element in decentralized systems.
• Ensuring trust through verification is vital for system stability.
• Dispute resolution is facilitated by verification mechanisms.
• Understanding verification’s role is crucial for decentralized systems’ success.
• Trust and verification are foundational to blockchain and AI integration.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.