10 Hours of Listening.
5 Minutes of Reading.

Deep dives into the conversations shaping the future of AI, Robotics & Crypto.

Save hours of your time each week with our key takeaways from the most interesting podcasts

🔍 Search & Filter
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

AI Podcasts

December 17, 2025

Rethinking Pre-Training for Agentic AI [Aakanksha Chowdhery] - 759

The TWIML AI Podcast with Sam Charrington

AI
Key Takeaways:
  1. Pre-Training is the New Frontier: The next leap in AI capabilities, particularly for agentic systems, will come from fundamental advancements in pre-training, not just post-training tweaks.
  2. Builders & Investors: Focus on teams rethinking loss objectives, curating high-quality reasoning data, and developing dynamic benchmarks for agentic capabilities. Be wary of "agentic" claims that lack foundational pre-training innovation.
  3. The "So What?": Over the next 6-12 months, expect a push for new benchmarks and data strategies that explicitly train models for multi-step planning, long-form reasoning, and error recovery, moving beyond simple next-token prediction.
See full notes
December 17, 2025

Code World Model: Building World Models for Computation – Jacob Kahn, FAIR Meta

AI Engineer

AI
Key Takeaways:
  1. Shift in AI Development: The focus moves from syntax-aware code generation to execution-aware reasoning, enabling more robust and intelligent code agents.
  2. Builder/Investor Note: Prioritize tools and platforms that support explicit execution modeling and highly asynchronous, high-throughput RL training for agentic systems.
  3. The "So What?": AI that can simulate complex systems internally will drastically reduce development and testing costs, accelerating innovation in software and distributed systems over the next 6-12 months.
See full notes
December 17, 2025

AI Kernel Generation: What's working, what's not, what's next – Natalie Serrino, Gimlet Labs

AI Engineer

AI
Key Takeaways:
  1. Strategic Shift: AI-driven kernel generation is not replacing human genius but augmenting it, allowing experts to focus on novel breakthroughs while AI automates the application of known optimizations across a complex hardware landscape.
  2. Builder/Investor Note: Focus on robust validation and hardware-in-the-loop systems. Claims of "AI inventing new algorithms" in this domain are premature. The real value is in automating the "bag of tricks" for heterogeneous compute.
  3. The "So What?": This technology is critical for scaling agentic AI workloads. Expect significant investment in tools that abstract hardware complexity and enable efficient, automated optimization, driving down the cost of AI inference in the next 6-12 months.
See full notes
December 17, 2025

Code World Model: Building World Models for Computation – Jacob Kahn, FAIR Meta

AI Engineer

AI
Key Takeaways:
  1. Semantic Shift: The future of AI in code moves from text generation to deep semantic understanding and execution simulation.
  2. Builder Opportunity: Develop next-generation debugging tools and code agents that leverage internal simulation for faster, more efficient development cycles.
  3. Investor Focus: Prioritize models and platforms that demonstrate explicit execution modeling, as this capability will redefine software development and create new market leaders.
See full notes
December 17, 2025

AI Kernel Generation: What's working, what's not, what's next – Natalie Serrino, Gimlet Labs

AI Engineer

AI
Key Takeaways:
  1. Infrastructure Shift: AI-driven kernel optimization addresses a critical bottleneck in scaling AI compute, enabling more efficient use of diverse hardware.
  2. Builder/Investor Note: Focus on solutions with robust, hardware-verified performance metrics and a clear human-in-the-loop strategy. AI is a powerful tool for automating optimization, not a magic bullet for novel algorithmic breakthroughs.
  3. The "So What?": This technology frees expert engineers from tedious optimization, allowing them to focus on higher-level research and truly innovative algorithmic design, accelerating the pace of AI development in the next 6-12 months.
See full notes
December 17, 2025

“How We Can Eliminate Crime” | Ben Horowitz and Garrett Langley

a16z

AI
Key Takeaways:
  1. Effective crime reduction requires a shift from reactive punishment to proactive, intelligence-driven deterrence, making it highly probable for criminals to be caught.
  2. The market for AI-powered public safety technology, particularly solutions that integrate data for precision and accountability, presents a significant opportunity. Public-private partnerships are a key funding mechanism.
  3. Over the next 6-12 months, expect to see more cities adopt advanced surveillance and AI tools, driven by private funding, as they seek to improve safety and address staffing shortages without resorting to ineffective, broad-stroke policies.
See full notes
December 17, 2025

“How We Can Eliminate Crime” | Ben Horowitz and Garrett Langley

a16z

AI
Key Takeaways:
  1. The Future of Policing is Intelligent: Integrating AI, drones, and smart cameras creates a precise, accountable, and safer policing model for both officers and communities.
  2. Invest in the "How": Builders and investors should focus on technologies that enhance certainty of capture, streamline judicial processes, and support public-private partnerships to modernize urban safety infrastructure.
  3. Safety Fuels Mobility: Eliminating crime is not just about law enforcement; it's about restoring the fundamental safety required for economic mobility and a functional society.
See full notes
December 16, 2025

Build reliable AI agents using W&B Training

Weights & Biases

AI
Key Takeaways:
  1. The democratization of RL fine-tuning will accelerate the development and deployment of more reliable and sophisticated AI agents across industries.
  2. Builders should explore open-source LLMs combined with RL fine-tuning as a cost-effective strategy to achieve specific performance benchmarks, especially where latency and cost are critical.
  3. Platforms abstracting infrastructure complexity and providing integrated tooling for the entire AI development lifecycle are crucial for the next phase of AI agent deployment.
See full notes
December 16, 2025

Build reliable AI agents using W&B Training

Weights & Biases

AI
Key Takeaways:
  1. The democratization of RL for LLMs will accelerate the deployment of more reliable and sophisticated AI agents across industries.
  2. Builders should move beyond basic prompt engineering and RAG. RL fine-tuning, now accessible via W&B Serverless RL, is a critical next step for high-stakes agentic applications.
  3. For the next 6-12 months, expect a surge in production-grade AI agents, with open-source models increasingly closing the performance gap with proprietary alternatives through advanced fine-tuning.
See full notes

Crypto Podcasts

August 18, 2025

One More Push Higher Before the Crash?

1000x Podcast

Crypto
Key Takeaways:
  1. The ETH Rally is an Illusion. Price action is dictated by treasury company flows, not fundamentals. Monitor their stock premium/discount to NAV as a leading indicator for the market top.
  2. Prepare for a "Stupid" Finale. The market is primed for one last FOMO-driven blow-off top. This is the signal to sell into strength, not add risk.
  3. Set Up the Next Home Run. The inevitable crash of treasury company stocks will present a massive opportunity. Prepare to buy these assets at deep discounts (30%+) to NAV when the market panics.
See full notes
August 18, 2025

What Liquid Funds Are Buying

Empire

Crypto
Key Takeaways:
  1. Concentrated Bets on Fundamentals Win. The era of "spray and pray" is over. The new meta is building highly concentrated portfolios (10-15 tokens) based on deep fundamental analysis of protocols with clear revenue models and product-market fit.
  2. Digital Asset Treasuries Are TradFi's On-Ramp. DATs are more than a short-term trade; they are the primary bridge for institutional capital to gain crypto exposure. Their marketing power is proving to be as crucial as their financial engineering.
  3. The 24/7 Market Is Coming. The tokenization of equities isn't a matter of *if* but *when*. This shift will create a fiduciary obligation for funds to move to on-chain assets, forcing a rapid, systemic evolution of financial markets.
See full notes
August 16, 2025

Why DATs Beyond BTC And ETH Are Risky

Empire

Crypto
Key Takeaways:
  1. **Concentrate on the Winners:** Bitcoin is the established store-of-value asset, and Ethereum is the dominant settlement layer for high-value digital assets. The data shows they have already won their respective categories.
  2. **The Rest is a Long Tail of Risk:** Investing outside of Bitcoin and Ethereum is a bet against powerful, gravity-like market forces. These alternatives are competing for a sliver of the market, increasing their risk of becoming obsolete.
  3. **Power Law is the Rule:** The market isn't about finding the "next" Ethereum; it's about recognizing that power laws are creating a duopoly where the vast majority of value will continue to accrue to the top two assets.
See full notes
August 16, 2025

Can ETH Keep Pumping Or Is It All Over?!

Steady Lads Podcast

Crypto
Key Takeaways:
  1. The New Game is Financial Engineering. The market's primary driver is the "Digital Asset Treasury" meta. Bitcoin leverages its "pristine collateral" narrative for debt financing, while Ethereum leverages native yield to justify its premium.
  2. Don't Expect a 2021 Redux. The institutional capital fueling this rally is not here to bid on your favorite altcoin. Their focus is on BTC, ETH, and treasury-related arbitrage, making a widespread, retail-driven altcoin season unlikely.
  3. De-Risk and Secure Profits. After a 3x run, seasoned traders are taking profits on ETH. The consensus is to refuse to round-trip your gains, pay down on-chain debt, and shift to scalping volatility rather than betting on a continued parabolic advance.
See full notes
August 15, 2025

The Best of Bell Curve

Bell Curve

Crypto
Key Takeaways:
  1. **Execution Guarantees Trump EVM Compatibility:** For complex financial products like derivatives, the ability to mathematically prove solvency outweighs the benefits of EVM compatibility, driving the rise of purpose-built L1s.
  2. **Memecoins Are a Macro Indicator:** Don't dismiss memecoins as a distraction. They are a direct, high-beta response to monetary debasement, signaling retail's desperation for returns in a broken financial system.
  3. **The Consumer War Is On:** While Ethereum solidifies its hold on institutional finance, the battle for consumer attention is just beginning. The success of its coordinated L2 strategy will determine if it can reclaim the narrative from chains like Solana.
See full notes
August 15, 2025

The Ether Machine Chairman: Real Risk of DATs, Who Wins From Stripe & Circle’s Chains | Roundup

Empire

Crypto
Key Takeaways:
  1. Structure Over Speed: In the DAT gold rush, avoid the shells. Reverse takeovers are fraught with hidden liabilities; cleaner de-novo SPACs are built for long-term institutional trust and better financing.
  2. Stick to the Winners: The DAT market will consolidate. Bet on pure-play vehicles for top-tier, liquid assets like ETH, as "Frankenstein" and illiquid-token DATs are destined for M&A or failure.
  3. Distribution is Destiny: In the payments war, Stripe’s direct ownership of millions of merchants gives it a crushing advantage over Circle’s middleware approach. Owning the customer is the only moat that matters.
See full notes