Unqualified AdviceUnqualified Advice
← All Episodes

Oct 5, 2026 Β· 53:09

Agency All the Way Down

Unqualified Fact Check

78%

held up

9Nailed it
7Close enough
0Whiffed it

Our vibe reading on our own accuracy: solidly meh, trending toward vibing.

Show Notes

What happens when AI makes building an idea almost free?

This week we introduce two projects born directly out of last week's podcast: Sean's Vibe Index, an attempt to quantify market and social sentiment from volatility, credit, meme stocks, prediction markets and Reddit; and Catsup Bot, our deeply unqualified experiment in buying the laggard of the Magnificent Seven.

But the bots lead somewhere bigger.

If AI can dramatically reduce the effort required to act on an idea, does it make us smarter? Or do we just stop thinking?

We talk about agency, cognitive abundance, the surprisingly small share of households paying for AI, the changing value of white-collar management, what an undergraduate education should actually teach, signs of consumer stress, household leverage, and a future where AI agents may disrupt the advertising-driven internet by shopping for us.

Plus: the Degeneracy Ratio, reinventing the associate's degree, whether β€œmeh” is bullish, and why sometimes you need to touch grass and make a pizza.

Find the Vibe Index and Ketchup Bot under Numbers at unqualifiedadvicepodcast.com.

Cheers, Sean & Dan

Programming Note: Off next week for fall break. We'll miss you!

Links

Timestamps

  • 00:00Cold Open and Welcome Back 00:34 - Revisiting the Ballpoint Pen (and the Slop Grenade) 03:44 - Introducing the Vibe Index 07:17 - Shiller CAPE at 41: Is Price Outrunning Earnings? 09:26 - ARK, Meme Heat and Betting on Belief 10:59 - Prediction Markets and the Degeneracy Ratio 14:32 - If We Hit Doomscrolling, Go Buy Something 18:45 - The Ketchup Bot: Buying the Most-Hated Mag 7 22:52 - How the Ketchup Bot Got Its Name 23:58 - Staying in the Saddle 29:01 - 98% of Households Aren't Paying for AI 33:59 - Finance and Tech Jobs Shrink: The Great Reshuffle 37:10 - Dan's Two-Year Undergrad 41:37 - Six Months From Hard Times? 45:49 - Agents, Walled Gardens and Who Pays for Ads 49:25 - Instinct and the One Percent Agent

The Fact Check, claim by claim

The Vibe Index says mehNailed it

Sean: The site reads 56, squarely in Meh, as of October 1 β€” which was a Thursday, not a Friday, but the vibes don't care what day it is.

A 2.1% credit spreadClose enough

Sean: It's HYG's return over IEF across the past quarter, a relative return rather than a yield spread β€” and Sean's correction to "actually pretty good" is right, at the 68th percentile.

Shiller CAPE at 41Close enough

Sean: 41.4 at Friday's close is right, but the December 1999 peak was 44.2, the long-run average is about 17, and Shiller's dataset is monthly, not daily.

Anthropic is about to IPONailed it

Sean: Anthropic confidentially filed a draft S-1 on June 1, 2026, and reporting has the float landing in October or November at $2 trillion or more.

What the vibes said nextNailed it

Sean: Doomscrolling 6 days and 100% up, Meh 76% up averaging +2.2%, Memes All the Way Down 20 days and 50% up, Nervous 54% β€” every number matches the site.

Bright Data, five bucks a monthClose enough

Sean: The 5,000 free requests a month are real, but there's no $5 plan β€” past the free tier it's pay-as-you-go at $1.50 per thousand.

The Ketchup Bot is up 200 bucksNailed it

Sean: Up $222.36 on $5,000 as of the October 2 close; the rule averages ranks rather than returns, but Tesla was dead last on all three lookbacks.

The ketchup came from a transcriptionNailed it

Sean: Last week's raw transcript renders Dan's buying the catch-up as you're constantly buying the ketchup β€” a Danism, laundered through Riverside.

98% of households aren't paying for AINailed it

Dan: Word for word from a16z's October 2 post, built on PNC data showing 2.2% of US households with a paid AI subscription.

Finance and tech are shedding jobsClose enough

Dan: We couldn't find the Warren Pies post behind the βˆ’246K / +812K figures, but the direction checks out β€” BLS shows financial activities down 129K from their May 2025 peak.

Jobs below expectations, but positiveNailed it

Dan: September payrolls came in at +29,000 against a consensus near 84,000 β€” positive, though July was revised to βˆ’10,000.

The Great ReshuffleNailed it

Sean: Jim O'Shaughnessy's phrase since 2021, for the once-in-generations shift from a physical, location-bound economy to a digital one.

Luxury is finally feeling the pressureClose enough

Dan: Real but not new β€” LVMH shares are down about 54% from their April 2023 high, and the reporting blames aspirational and Chinese shoppers more than the wealthy.

Five grandNailed it

Dan: Trump pledged $5,000 to every adult citizen on September 9 if Republicans keep both chambers, and repeated it October 2–3.

Google isn't a monopolyClose enough

Dan: A respectable aggregation-theory argument, but a federal judge ruled Google an illegal search monopolist in August 2024, and the ruling stands on appeal.

Instinct takes one percentClose enough

Dan: Free, with a cut-of-the-purchase model, but no take rate has been published; it just raised $1 billion at a $10 billion valuation.

Predictions from this episode

PendingDaniel (Sean poses the "if this is true, in six months what happens?" frame)

Conditional. *If* group psychology is about six months from having "endured the right amount," then by roughly April 2027 the US prints negative payroll numbers repeatedly and diesel goes above $7/gal, which feeds angst into everything else and pushes politics toward relief checks.

Said in: Agency All the Way Down (Oct 4, 2026)Horizon: ~6 months (to 2027-04-04). Evaluable on BLS headline payrolls (two or more negative first prints, or negative after revision, between Oct 2026 and Apr 2027) and on EIA's weekly US on-highway diesel average (any print above $7.00).Target: Dec 31, 2027Confidence: Exploratory and explicitly conditional. Dan was thinking aloud ("I don't know"), not committing. Score each leg on its own. Context at recording: September 2026 payrolls came in at +29K against ~84K expected, and July was revised to βˆ’10K, so one negative month is already on the books. Unemployment was 4.2%. EIA diesel set an all-time weekly record of $6.529 on 2026-09-21 (well past June 2022's $5.810) and stood at $6.382 on 2026-09-28, so "$7" is about 10% away. Pairs with `more-inflation-ahead-2026-2027` and `september-2026-economic-sting`. On air it sits beside Dan's "I think what I mistaked was… we could cut rates… but it's a supply based constraint."
PendingSean & Daniel (Sean names the management layer and endorses the "intensify and expand" call; Dan supplies the coordination-vs-decision distinction)

The AI-era job reshuffle hits the coordination/management layer hardest, while people who make good decisions, lead and align teams "float upward." The finance+tech job losses already visible intensify in 2027 and spread to other white-collar industries.

Said in: Agency All the Way Down (Oct 4, 2026)Horizon: 2027 for the "intensify and expand" leg; ~5 years for the layer-level claim. Evaluable on BLS CES employment for financial activities and information through 2027 vs. total private, and on whether professional & business services (management/administrative support) turns negative year over year.Target: Dec 31, 2027Confidence: Load-bearing for the show's running atom-economy thesis. A narrower, more measurable cousin of `upper-middle-class-gets-walloped`. Context at recording: the Pies figures could not be located (X blocks fetching), but BLS shows financial activities βˆ’129K from its May 2025 peak (mostly insurance) and βˆ’7K in September 2026. Bloomberg (Jul 2026) put tech+finance losses at ~28K a month through May 2026. Note the tension with `layoffs-were-overhiring-not-ai` (Ep 52): the jobs data can't tell overhiring from AI displacement, so score the *pattern*, not the cause.
PendingDaniel vs. Sean β€” logged as one contested entry

Two-sided. Daniel: if AI agents pull advertising dollars out of marketplaces, consumers end up paying *more*. Platforms need new revenue, customer matching gets harder, and sellers must go expensively multi-channel, so total costs rise and winners keep winning. Sean: sellers' ad bills fall, so prices could fall, or more likely sellers simply keep the savings as margin.

Said in: Agency All the Way Down (Oct 4, 2026)Horizon: 5–10 years, and only evaluable if agentic shopping takes a material share of e-commerce. Watch retail-media ad revenue (Amazon, Walmart Connect, Instacart, DoorDash) vs. agent-mediated GMV, and seller take rates and margins.Confidence: Genuinely contested and mechanism-rich. Sean's own closing line, that the effects are all "Siamese twins," concedes it may never be cleanly scoreable. The ad-dependence premise came from Ep 59 (DoorDash, Instacart and Amazon: "basically every profit dollar for the marketplace is advertising"). The live business-model counterexample on air was Instinct: free, monetized by a cut of purchases, $1B Series C at $10B (2026-09-28). Its take rate is unpublished, not the "one percent" said on air. Pairs with `walmart-more-ai-native-than-amazon`, `meta-over-google-in-the-agent-decade` and `agent-layer-becomes-the-os`.
PendingSean (Daniel supplies the ad-dependence mechanism and says he is already shifting his own selling effort)

Walmart becomes the agent-friendly retailer and Amazon the agent-hostile one. Because Walmart's economics depend far less on ad revenue, it can open its catalog to personal agents, and sellers shift optimization effort toward Walmart as a result.

Said in: Where Are the Humans? (Sep 27, 2026)Horizon: 2–3 years. Evaluable on whether Walmart's agent partnerships widen while Amazon keeps blocking third-party agents, and on Walmart Marketplace seller growth relative to Amazon's third-party seller growth.Updated: Oct 4, 2026Confidence: Load-bearing, and already half-landed before it was said: Walmart joined Muse as a launch shopping partner at Meta Connect on Sep 23, four days before recording, while Amazon had blocked Muse days earlier (and has previously blocked Google's and OpenAI's shopping agents). The open, scoreable part is durability. Amazon also runs its own shopping agent, and the Tech Times framing ("standards it ignores for its own shopping agent") suggests a closed-garden strategy, not a refusal to do agents. The direct heir of `walmart-as-amazon-challenger` (pilot, 2024), which was scored Partially on its pro-US-SMB framing. This version drops that framing and bets on ad-dependence instead.
PendingDaniel (Sean calls it "a possibility" and asks about durability)

Over the next decade the comfortable upper middle class sees the weakest earnings and net-worth trajectory of any cohort, as high earnings built on credentialed white-collar work stop compounding. Trades, manufacturing and other atom-economy "doers" move up, and the resulting status anxiety becomes a political force.

β€œMy pet theory is that the upper middle class is the one that's gonna get walloped over the next decade… their expenses are too high because they have high earnings. And I'm questioning their ability to have high earnings going forward the way history has foretold." (The political half): "It's the two percent where revolutions come from. I think it's the upper middle class that is where the angst is… the near to power who are watching themselves and their cohort backslide." (The other side of the trade): "The tradespeople, the manufacturing, the doers, the people in the physical and atoms world, they're gonna come up.”
Said in: Where Are the Humans? (Sep 27, 2026)Horizon: 10 years. Evaluable against Fed Distributional Financial Accounts net-worth growth by percentile (the 50th–90th and 90th–99th bands versus the bottom 50% and top 0.1%), and against trades-vs.-professional wage growth.Updated: Oct 4, 2026Confidence: Load-bearing for Dan's standing worldview. It is the explicit bridge between `atom-economy` (Ep 47), `tacit-knowledge-edge` (Ep 48) and the Ep 56 "revolutions come from the top two percent" beat, which he calls back here. His supporting claim, that the bottom 50%, the deciles above it and the top 0.1% have seen the largest net-worth gains while the upper middle has seen the smallest, was not scored in this episode's fact-check and should be checked against DFA before it gets repeated. Current shares (Q2 2026 DFA): top 0.1% 15.0%, 90th–99th 36.4%, 50th–90th 28.8%, bottom 50% 2.3%. Sean's durability question ("is this a trend that's durable over a five- or ten-year period?") is left open; Dan's answer is that it's "up to the hive mind."
PendingSean (made-with vs. made-by) vs. Daniel ("until it becomes water") β€” logged as one contested entry

Two-sided. Sean: public perception of AI-assisted work settles on a distinction between things made *by* AI (slop, judged harshly) and things made *with* AI (human intent visible, judged as craft). Daniel: the distinction dissolves instead. Like the phonograph, AI use "becomes water," watermarking fails because non-US models won't comply, and nobody asks who held the pen.

Said in: Where Are the Humans? (Sep 27, 2026)Horizon: 5–10 years. Scoreable on whether a durable "made with AI" label or norm emerges in creative markets (awards eligibility, platform labeling, the artist community Sean describes), or whether disclosure norms fade the way "digitally recorded" did.Updated: Oct 4, 2026Confidence: The episode's intellectual centerpiece and its working title, set off by Sean reading Vonnegut's blue-footed booby passage from *GalΓ‘pagos*. Genuinely contested, and each host concedes the other's strongest point: Sean accepts that it "becomes water" eventually, and Dan accepts courtroom evidence as the case where provenance matters. Both positions can end up true in sequence (a made-with norm first, then dissolution), so score which one describes 2030, not which was right in principle. Sean's digital-vs-film analogy is his own best evidence for Dan's side. Pairs with `ai-creativity-jobs-fears-unfounded` and Ep 57's `comedy-before-science-as-agi-marker`.