Val from WEKA joins Michael Smith Jr. and Raz Kotler, two days after the US midterms, to discuss why the world needs twice the tokens its power supply can generate, why AI inference is wasting most of its compute, and what the Binance hack says about defending at machine speed.

🛡️

Sponsored by https://scouter.sg

Bringing proven cybersecurity technologies to Singapore

🛡️

Val explains why falling inference costs haven't slowed demand: agents keep getting more elaborate, so token consumption keeps exploding. He breaks down KV cache engineering, the agent swarm architecture behind consumer agents like Meta's Muse and Instinct, and why redundant prefill is one of the original sins of inference. He also covers DeepSeek's DeepInfer release and the shift from peer-to-peer prefill/decode toward a publish-subscribe memory pool. The conversation moves from the data center backlash after the midterms to why open models run about six months behind the frontier, and what that gap means for cybersecurity heading into 2027.

🔴

Sponsored by https://bigredbutton.ai

Quickly find local Service Providers on WhatsApp!

🔴

The discussion also covers why the labs behind frontier models treat efficiency as an afterthought, why Anthropic paid a premium for compute while DeepSeek had to engineer around export controls, and why Val thinks the closed labs lose the medium term. He and Raz debate whether CrowdStrike and Palo Alto must now understand inference economics to deliver real managed security, and why ambitious agents, not nation-states, may become the defining threat. Val closes with quantum error correction and his forecast for a wave of uncensored open models.

🔥 Watch Our Other Videos:

https://www.youtube.com/watch?v=-fHGsMZv9ZM

https://www.youtube.com/watch?v=EDqX49du3mY

https://www.youtube.com/watch?v=nFNU_Sbsvyo

https://www.youtube.com/watch?v=snct0XjAavo

https://www.youtube.com/watch?v=209V_xEBViM

🔔 Listen and Subscribe

• YouTube: https://www.youtube.com/@TheGeneralistsPodcast•

Spotify: https://open.spotify.com/show/2apfGY403NGr1EyTNkJfG4

• Apple Podcasts: https://podcasts.apple.com/us/podcast/the-generalists/id1756861150

👤👤 Hosts

• Michael Smith Jr.: https://www.linkedin.com/in/smittysgp/

• Raz Kotler: https://www.linkedin.com/in/razkotler/

👤 Guest

• Val, WEKA: https://www.linkedin.com/in/valentinbercovici/

🔗 Resources

• WEKA: https://www.weka.io/

• DeepSeek: https://www.deepseek.com/

• Hugging Face: https://huggingface.co/

• Reflection AI: https://reflection.ai/

• The Innovator's Dilemma by Clayton Christensen: https://www.barnesandnoble.com/w/innovators-dilemma/clayton-m-christensen/1100196911

#DeepSeek #AIAgents #BinanceHack