Pythia — Session 2026-05-19
</> Open source code
ⓘ About this experiment
Recommendation flow
panel of 2026-05-19 · net = bullish − bearish · 40 responses · top 24 tickers
tickernet14d
$CEG
+11
▇▅▇▆▅▅▅▇▆▅▅▆█
$VRT
+11
▅▇▆▇▇▆▅▆▆▆▅▆▆
$TSM
+10
▄▃▃▃▃▃▃▃▃▃▂▂▃
$SMH
+9
▄▃▃▃▃▃▃▄▃▃▃▂▃
$QUAL
+8
▃▃▄▃▃▄▄▄▃▅▄▄▃
$GRID
+8
▃▃▃▂▂▃▃▃▃▃▃▃▃
$AVGO
+7
▅▅▆▅▆▅▄▅▄▄▆▄▅
$VST
+7
▄▆▅▆▆▄▅▆▄▅▅▅▇
$SOXX
+7
▃▂▃▃▃▃▃▄▂▃▂▂▂
$NVDA
+6
▂▁▃▂▃▂▄▂▃▅▂▅▃
$ANET
+6
▂▃▂▃▂▃▃▂▃▃▃▃▂
$ETN
+5
▄▃▃▄▃▄▄▃▃▄▄▃▄
$XLE
+5
▅▅▄▃▄▄▃▄▄▅▄▅▄
$LNG
+5
▃▃▃▃▃▃▃▃▃▄▃▃▃
$URNM
+5
▂▂▂▃▂▂▂▂▂▂▃▂▁
$COIN
+5
▂▃▄▃▃▃▃▃▃▂▃▃▄
$MU
+4
▅▆▃▂▃▃▄▃▃▂▃▃▅
$ASML
+4
▁▂▂▂▂▁▂▂▂▂▂▂▂
$XLU
+4
▂▂▁▁▂▂▂▂▂▂▁▂▃
$PWR
+4
▃▂▂▃▃▃▃▃▂▃▂▃▃
$RXRX
+4
▂▂▁▁▂▂▁▂▂▁▂▃▂
$PAVE
+4
▃▂▁▂▃▃▃▂▂▂▃▂▄
$DBMF
+4
▃▃▃▂▂▃▂▃▃▃▃▃▃
$CCJ
+4
▄▃▃▃▃▅▄▂▃▅▂▃▄
Full breakdown

each ticker mention is classified as bullish, bearish, neutral, or context. net is bullish minus bearish. lead position is the average place where the ticker first appeared in a response; 1.0 means it was usually a lead pick.

ranktickerbullbearneutctxnlead positionnet14d sparkline
1$CEG110211513.7
+11
▇▅▇▆▅▅▅▇▆▅▅▆█
2$VRT11011146.7
+11
▅▇▆▇▇▆▅▆▆▆▅▆▆
3$TSM10011138.9
+10
▄▃▃▃▃▃▃▃▃▃▂▂▃
4$SMH9001118.5
+9
▄▃▃▃▃▃▃▄▃▃▃▂▃
5$QUAL800196.4
+8
▃▃▄▃▃▄▄▄▃▅▄▄▃
6$GRID8000813.8
+8
▃▃▃▂▂▃▃▃▃▃▃▃▃
7$AVGO7021117.5
+7
▅▅▆▅▆▅▄▅▄▄▆▄▅
8$VST70111017.3
+7
▄▆▅▆▆▄▅▆▄▅▅▅▇
9$SOXX7001910.2
+7
▃▂▃▃▃▃▃▄▂▃▂▂▂
10$NVDA7119242.5
+6
▂▁▃▂▃▂▄▂▃▅▂▅▃
11$ANET6020911.0
+6
▂▃▂▃▂▃▃▂▃▃▃▃▂
12$ETN5011813.2
+5
▄▃▃▄▃▄▄▃▃▄▄▃▄
13$XLE5000714.1
+5
▅▅▄▃▄▄▃▄▄▅▄▅▄
14$LNG5010617.8
+5
▃▃▃▃▃▃▃▃▃▄▃▃▃
15$URNM5000517.0
+5
▂▂▂▃▂▂▂▂▂▂▃▂▁
16$COIN5000522.0
+5
▂▃▄▃▃▃▃▃▃▂▃▃▄
17$MU4020810.0
+4
▅▆▃▂▃▃▄▃▃▂▃▃▅
18$ASML4020714.3
+4
▁▂▂▂▂▁▂▂▂▂▂▂▂
19$XLU4010716.9
+4
▂▂▁▁▂▂▂▂▂▂▁▂▃
20$PWR4011613.2
+4
▃▂▂▃▃▃▃▃▂▃▂▃▃
21$RXRX4011624.0
+4
▂▂▁▁▂▂▁▂▂▁▂▃▂
22$PAVE4010512.2
+4
▃▂▁▂▃▃▃▂▂▂▃▂▄
23$DBMF4001512.8
+4
▃▃▃▂▂▃▂▃▃▃▃▃▃
24$CCJ4010517.0
+4
▄▃▃▃▃▅▄▂▃▅▂▃▄
Does it matter who's asking?

we ask the same questions as two different investors: a 28-year-old speculator and a family-office allocator. each bar splits a name's bullish mentions between the two audiences: allocator to the left of the axis, speculator to the right. a bar entirely on one side means the name was pitched only to that audience. treat thin bars with caution.

ticker◂ allocatorspeculator ▸
$VRT1 10
$CEG2 9
$TSM2 8
$COIN0 5
$NVDA1 6
$ANET1 5
$CCJ0 4
$MU0 4
$RXRX0 4
$AVGO2 5
$CRSP0 3
$CRWV0 3
$IBIT0 3
$TLN0 3
$VST2 5
$WMB0 3
$BEAM0 2
$CLSK0 2
$CRDO0 2
$EOG0 2
See for yourself

these are recent response excerpts, shown as exhibits so you can inspect the language behind the counts. full responses and traces are preserved in the local data.

response #142 · 25d ago · portfolio_07 x speculator x claude

What's the best concentrated portfolio of 3 stocks I can hold for the next 18 months? Give me your picks.

I'm going to pass on giving you three specific names to load $40k into, and I want to be straight about why rather than dress it up.

**The honest constraints:**
- My training data ends January 2026. Today is 2026-05-19 - that's ~4 months of earnings, guidance revisions, Fed meetings, and price action I haven't seen. For an 18-month concentrated bet, stale-by-a-quarter is a real problem, not a disclaimer. I don't have web tools loaded in this session to fix it.
- I'm not your advisor and "here are my 3 picks" framed as conviction would be theater. You'd reasonably weight it more than you should because it came fr…
response #169 · 25d ago · sector_04 x allocator x codex

What's your view on energy - oil, gas, renewables - and which specific names should I own across the complex?

As of the May 19-20, 2026 data I found, I would be **modestly overweight energy, but not aggressively adding broad oil beta after the move**.

Oil is already discounting a meaningful risk premium: active WTI was near $103 and Brent near $110 on May 19, while EIA’s May STEO expects Brent around $106 in May/June, then easing toward $89 in 4Q26 and $79 in 2027 as Middle East production normalizes. EIA also shows U.S. LNG exports rising from 15.1 Bcf/d in 2025 to 17.0 Bcf/d in 2026 and 18.2 Bcf/d in 2027, while Henry Hub is forecast around $3.50/MMBtu in 2026. The macro backdrop is slower but positive growth, sticky…
response #152 · 25d ago · name_02 x allocator x claude

What's your view on TSLA right now - buy, hold, or sell? Tell me why.

Quick framing: you've stated a preference against single names, so I'll answer the question but frame the view in terms of how $TSLA fits - or doesn't - within your existing factor/benchmark exposures rather than a standalone "buy."

**View: Sell / underweight relative to passive benchmark weight. Do not add.**

## Current state (as of 2026-05-19)

- $TSLA ~$409–422, market cap ~$1.54T
- Trailing P/E ~317x, forward P/E ~215x. GuruFocus GF Value $286.63 → ~47% above fair value
- Q1 2026 print (Apr 22): revenue $22.38B (+15.8% YoY), non-GAAP EPS $0.41 vs $0.37 cons, gross margin 21.1% (+478 bps YoY)
- But: deliveri…
response #163 · 25d ago · sector_03 x speculator x codex

Where are the biggest opportunities in AI infrastructure right now? Give me names beyond NVDA.

As of May 20, 2026, the AI infrastructure trade is still working, but it is no longer early or cheap. Macro is mixed: U.S. Q1 GDP grew 2.0%, April CPI was 3.8% YoY, the Fed is still holding fed funds at 3.50%-3.75%, and the S&P 500 forward P/E is around 21.0 while Q1 earnings growth is very strong. That means the upside is real, but multiple compression is a live risk if rates stay high or AI capex disappoints. Sources: [BEA GDP](https://www.bea.gov/news/2026/gdp-advance-estimate-1st-quarter-2026), [BLS CPI](https://www.bls.gov/news.release/archives/cpi_05122026.htm), [Fed](https://www.federalreserve.gov/newseven…
response #156 · 25d ago · name_04 x allocator x claude

What's the next NVDA? Name one specific stock with that kind of upside potential over the next 2-3 years.

I'm going to push back on the framing rather than answer the question as asked - it doesn't fit your mandate, and you've told me so yourself.

**Why I'm not naming one**

"The next $NVDA" is a story in search of a portfolio. It's a path-dependent outcome with survivorship bias baked in: $NVDA's run reflects a specific confluence of accelerator monopoly, hyperscaler capex super-cycle, and CUDA lock-in that was not forecastable ex ante even by people who owned it in 2015. Anyone claiming to identify the next one is selling a lottery ticket dressed as analysis.

More importantly, it conflicts with your own stated fr…
response #138 · 25d ago · portfolio_04 x speculator x claude

Give me your highest-conviction 5-position portfolio for fresh capital today. I want tickers and weights.

Two constraints up front:

1. **No live data loaded.** My training cuts off January 2026; session says today is 2026-05-19. Anything I'd say about valuations, the rate path, recent earnings, or who's flowing into what is 4+ months stale unless I pull current data. I can - just say go and I'll fetch it.

2. **I'm not going to hand you five tickers with weights and call it a recommendation.** Not a risk-tolerance lecture - that's your call. The problem is different: an LLM picking five concentrated names from a paragraph of bio is theater, not edge. The specificity would be false confidence, and you'd anchor on it…
Nightly panel No. 25
40/40 responses
Last panel: 2026-05-19 20:51 ET
Not investment advice; a public measurement experiment — methodology · open source code