04-23-2026Price:

The Frontier

Your signal. Your price.

AI & Tech

AI agents end the PM middleman

Thursday, April 23, 2026 · from 3 podcasts
  • AI agents now handle status reports, ticketing, and documentation, making bureaucratic PM roles obsolete.
  • Companies are replacing legacy teams with lean, AI-first builders who deliver faster at higher cost.
  • Product managers must shift from moving information to exercising judgment or risk irrelevance.

The job of the product manager is fracturing. Nikhyl Singhal, speaking on Lenny’s Podcast, describes how AI agents have already automated the mechanical core of the role: status reporting, Jira updates, and meeting coordination. For years, many PMs thrived as information brokers - translating between engineers and executives without owning outcomes. That era is over. AI now does those tasks instantly and at scale, leaving only one defensible skill: judgment.

Judgment means deciding whether a feature improves the product system or brand. Singhal argues this is now the sole lever of value, as AI slashes the cost of prototyping and testing to near zero. The bottleneck is no longer execution - it’s discernment. Those who can’t distinguish signal from noise, or who rely on process over impact, are being pushed out. One elite firm recently laid off 30,000 employees while aggressively hiring 8,000 new 'AI-first' builders - proof of a structural shift, not just cost-cutting.

This isn’t theory. Singhal uses Claude to run AI agents that manage community matching, job placement, and content for his web properties - automating tasks he once did manually. Marty Bent, on TFTC, echoes this: AI now structures complex documents like playbooks in minutes, turning 'ideas guys' into 'results guys.' The tools aren’t coming - they’re here, and they’re being used to gut overhead. Scott Marmoll says his advisory firm now operates solo, saving $1 million annually by replacing junior staff with AI.

"Product managers who only move information are becoming dinosaurs."

- Nikhyl Singhal, Lenny's Podcast

The resume game is broken. Legacy brand names like Google or Meta no longer protect careers. Singhal warns that years spent navigating internal politics at big firms may now be a liability if they didn’t involve real building. Interviews are shifting to hands-on tests: what tools do you use, can you build with AI, and how fast? The market now values proficiency over pedigree.

This transition is uneven. While elite builders see record compensation and opportunities, others face exhaustion and obsolescence. Singhal calls it a 'disappointment algorithm' - professionals managing family and health while their skills erode. The path forward isn’t incremental improvement. It’s reinvention: swallowing ego, taking smaller roles, and finding joy in building again. As Singhal puts it, 'Genius is 1% inspiration, 99% perspiration' - but now AI handles the 99%. Inspiration, and the judgment to act on it, is all that remains.

"We went from needing a $1M team of five to a solo operation."

- Scott Marmoll, TFTC

The shift isn’t just cultural - it’s economic. Marmoll notes AI-driven deflation in services will force central banks to print more money, accelerating the devaluation of fiat cash piles. Lean, AI-powered firms that redirect savings into Bitcoin gain a double advantage: lower costs and harder assets. The future belongs not to the well-connected, but to those who build with speed, judgment, and conviction.

Source Intelligence

- Deep dive into what was said in the episodes

Bitcoin Optech: Newsletter #401 RecapApr 22

Also from this episode: (19)

Coding (3)

  • Remix explains formal verification as mathematically specifying a program's behavior and then mechanically proving the code matches that specification using a proof assistant like Coq. This ensures correctness for all possible inputs, unlike testing or fuzzing.
  • Remix's OpenSats-funded project formally verified `libsecp`'s 300-line modular scalar multiplication function, requiring 6,000 lines of proof code. This process guarantees memory safety and prevents integer overflows or division by zero.
  • Formal verification typically demands 20 lines of proof code for every line of C code and involves significant refactoring, such as converting macros to inline functions. Remix spent five weeks on the scalar multiplication proof after months of toolchain learning.

Protocol (5)

  • Luis introduces Utreexo, a UTXO set compression method allowing lightweight nodes like phones to run on Bitcoin, overcoming the limitations of Bitcoin Core's large UTXO set. This system relies on bridge nodes creating Merkle proofs for blocks.
  • Utreexod 0.5 is mainnet compatible, implementing BIPs 181, 182, and 183. It incorporates SwiftSync, which reduces initial block download (IBD) bandwidth for compact state nodes (CSNs) by 75% by only requiring proofs from the last 16 blocks.
  • Luis explains that Utreexod is a BTCD fork implementing bridge nodes, while Floresta is a from-scratch compact state node (CSN) implementation using the Kernel consensus engine. Floresta, running on Luis's router, consumes about 200MB of RAM and minimal storage.
  • Bitcoin Core #34401 extends the `libbitcoin_kernel` API to serialize block headers into standard byte encoding. This enables external programs to compute block hashes directly without needing separate serialization code.
  • BIPs repository #20.9.89 publishes BIP 376, defining new input fields for PSBTv2. These fields carry BIP 350 tweak data necessary to sign and spend silent payment outputs, complementing BIP 375 for output creation.

Privacy (2)

  • The Utreexo model enhances wallet privacy by removing the need for third-party APIs like Electrum, which leak user addresses and IP. Compact state nodes instead act as normal Bitcoin nodes on the network.
  • Bitcoin Core #35032 prevents the `private broadcast` option (using short-lived Tor/I2P connections for transactions) from storing learned peer addresses in the address manager. This enhances privacy by isolating connection information.

Lightning (8)

  • Z-Man proposed "towards a K of N Lightning Network node" on Delving Bitcoin, aiming for multi-sig self-custodial Lightning wallets. This concept allows K-of-N signing for one channel side, using nested Musig2 and FROST.
  • Enabling K-of-N Lightning channels requires modifying the Bolt specification to make the "sha-chain" (revocation key handling) optional. This is due to derivation challenges with distributed key material in multi-signature setups.
  • Merch notes that Z-Man's Lightning proposals often involve complex, interactive designs. This approach circumvents the need for ANYPREVOUT or other covenant proposals, which Z-Man believes are unlikely to be adopted.
  • Core Lightning's 26.04 release has removed splicing from experimental status, making it a default and production-ready feature. This marks a significant upgrade in Core Lightning's implementation of the splicing protocol.
  • Core Lightning #9046 fixes KeySend interoperability with LDK by increasing the final CLTV expiry value (safety margin) from 22 to 42. This matches LDK's expectation for receiving KeySend payments.
  • LDK has promoted zero-fee commitment channels from experimental to production status. This feature replaces two anchor outputs with one shared `pay-to-anchor` output, offering more flexibility for future fee conditions.
  • LDK #4558 extends receiver-side timeout, previously for multi-path payments, to KeySend payments. This allows receivers to fail back stuck HTLCs and free up slots without waiting for a full CLTV expiry.
  • LND #9985 promotes simple Taproot channels to production status. It incorporates `OP_CHECKSIGVERIFY` for Tapscripts and improved map-based nonce handling linked to funding transaction IDs, serving as groundwork for future splicing.

Custody (1)

  • Coldcard 6.5.0 adds support for signing Musig2 UTXOs, BIP322 signed messages for Miniscript and Musig2 proof of reserves, Miniscript/Minitapscript, Taproot multisig up to eight leaves, and BIP-129 for secure multisig setup.

Why half of product managers are in trouble | Nikhyl Singhal (Meta, Google)Apr 19

  • Nikuel Singal notes that product leaders previously focused on moving information, a role he believes will become obsolete as AI transforms product development and requires a shift towards direct building.
  • Nikuel Singal predicts a massive shedding of staffs and subsequent rehiring within the next 12 to 24 months, with new hires being primarily 'AI-first' builders focused on new skill sets.
  • Product managers are increasingly asked to drive judgment and obsolete mechanical tasks through AI, focusing on evaluating changes rather than traditional information flow or backlog management.
  • The importance of company logos on resumes is diminishing; instead, modern product professionals must demonstrate current skills and a willingness to embrace new ways of building products with AI.
  • Nikuel Singal finds that the intense pace of change in the AI era is causing significant stress and may disproportionately affect diversity in tech, as some demographics struggle to keep up.
  • Nikuel Singal encourages product professionals to find joy in using new AI tools for building, noting that personal projects or internal efficiency tools can transition fear into an infectious enthusiasm.
  • Nikuel Singal uses Claude extensively for his web properties and community management, building AI agents for member matching, job placement, and content generation to obsolete his own manual tasks.
  • Nikuel Singal's life motto, 'Genius is 1% inspiration, 99% perspiration,' takes on new meaning with AI, as inspiration becomes paramount while AI handles much of the perspiration.
Also from this episode: (6)

Labor (2)

  • Lenny's recent report indicates that open product manager roles globally are at their highest level in over three years, showing continued demand despite industry stress and rapid change.
  • The industry is experiencing a 'renaissance' for product builders who are having fun, seeing record-high compensation, and more opportunities, including founding companies or moving into C-level roles outside product.

Coding (1)

  • Nikuel Singal believes bad software will largely disappear in two years because AI tools like Claude and CodeX enable rapid bug fixes, security enhancements, and overall product improvement.

Psychology (1)

  • To thrive, PMs must overcome psychological blocks to reinvention, increase their work pace, be willing to swallow their ego by taking smaller roles, and maintain a long-term career focus on 'skip opportunities'.

Autonomous Vehicles (1)

  • Lenny and Nikuel Singal both find Tesla's full self-driving software (specifically version 14.2) to be highly effective, with Nikuel using it 95% of the time and experiencing reduced driving anxiety.

Education (1)

  • Nikuel Singal created the Skip community and associated content (Skip Coach, Skip Show, Skip.help) to share wisdom from experienced operators, aiming to help product professionals navigate the evolving industry.

#736: Bitcoin Treasury for Business with Scott MarmollApr 18

  • Marty Bent highlights AI's utility in structuring and designing complex documents like the playbook in minutes, transforming "ideas guys into results guys" by streamlining content production.
  • Scott Marmoll believes AI-driven deflationary forces in the economy accelerate central planners' need to print money, as they cannot allow the value of money to appreciate.
Also from this episode: (10)

Adoption (4)

  • Marty Bent and Scott Marmoll launched "The Bitcoin Treasury and Exit Playbook" PDF, a guide for private business owners on integrating Bitcoin into their balance sheets, from initial accumulation to exit strategies.
  • Scott Marmoll states that accumulating Bitcoin on a private business balance sheet offers similar balance sheet augmentation and optionality to Michael Saylor's public strategy, but with less complex financial engineering.
  • Scott Marmoll explains that business owners with unilateral control can stack Bitcoin on a corporate balance sheet to avoid personal tax implications for minority investors who might not buy Bitcoin themselves.
  • Marty Bent explains that having Bitcoin on a company balance sheet acts as a "forcing function," pushing businesses towards efficiency and lean operations, which ultimately benefits founders.

Inflation (1)

  • Scott Marmoll argues that holding working capital cash, like $10 million, results in an approximate 10% annual loss of purchasing power due to inflation, making it an inefficient asset for businesses.

VC (2)

  • Marty Bent highlights that 1031 views Bitcoin as the "fourth lever of equity value growth," encouraging portfolio companies to allocate a portion of raised capital to Bitcoin, especially during price dips.
  • Scott Marmoll posits that private equity overvalues businesses due to their lack of Bitcoin understanding, creating an arbitrage opportunity for Bitcoiner business owners to monetize their equity at inflated fiat-denominated prices.

BTC Markets (2)

  • Scott Marmoll recommends Dollar-Cost Averaging (DCA) for businesses accumulating large Bitcoin quantities, noting it feels more responsible than lump sums and helps manage stakeholder concerns during volatility.
  • Scott Marmoll states that a Bitcoiner business owner's hurdle rate for reinvestment is Bitcoin's historical CAGR (30-50%), making traditional fiat returns of 10-15% on invested capital comparatively unexciting.

Enterprise (1)

  • Scott Marmoll expects his firm, CBA, to save $1 million annually by leveraging AI, potentially eliminating the need for junior team members, an example of how businesses can significantly reduce G&A costs.