🏛️ The AI trust-and-safety layer is the next censorship war
Consumer AI products are not raw models: a human-programmed trust-and-safety layer intercepts the prompt and decides whether the model's real answer ships. Sacks' claim is that this layer carries the monoculture bias of a handful of companies, that they cannot see it and will not hand users a filter, and that it is a godlike power to reprogram what people learn - so the model providers inherit social media's censorship war as a reputational and eventual regulatory overhang. Friedberg and Chamath argue it is a routine media editorialization cycle that resolves itself as models fragment by audience.
0
CONVICTION
peaked 57.8 WATCH
EMBER
band · contested
▼ SHORT MSFT
expression · bearish politics-market
MISS
outcome · R -57.9% · α -82.4%
2024-02-17
window closed
⚠ CONFLICTING IDEAS ON THE SAME TICKER
— the board is arguing with itself; net it before trading (ticker view →)
MSFT ·
SaaSpocalypse is overdone for compliance-moat enterprise software
LONG
GREEN 71.0
← stronger than this idea (23.3)
GOOGL ·
Anthropic IPO (~Oct 2026) is attractively priced at $1.5-2T
LONG
WATCH 53.9
← stronger than this idea (23.3)
GOOGL ·
Personal AI agents: Google's data moat wins the form factor
LONG
EMBER 29.3
← stronger than this idea (23.3)
MSFT ·
Frontier AI labs keep durable premium pricing (not commoditized)
LONG
EMBER 26.6
← stronger than this idea (23.3)
GOOGL ·
Frontier AI labs keep durable premium pricing (not commoditized)
LONG
EMBER 26.6
← stronger than this idea (23.3)
MSFT ·
Own AWS, Azure and GCP and nothing else — the clouds capture the AI application dollars
LONG
EMBER 20.6
GOOGL ·
Own AWS, Azure and GCP and nothing else — the clouds capture the AI application dollars
LONG
EMBER 20.6
also touching these tickers, same direction: Blue cities ban or license-cap robotaxis 28 (GOOGL) · Tort floodgates open on social media 15 (GOOGL) · ChatGPT takes Google search share and the market starts pricing the decay 9 (MSFT) · ChatGPT takes Google search share and the market starts pricing the decay 9 (GOOGL)
⚖ Why this verdict
fully deterministic — evaluate.py replays this from daily closes; nothing below is editable or hand-set
- Window: 2023-02-17 → 2024-02-17 — first mention + 12-month horizon, then the window locks.
- The call: ▼ SHORT MSFT (primary play). MSFT rose 57.9% over the window → direction-adjusted R = -57.9% (the call lost money).
- Benchmark: holding SPY over the same window returned +24.5%
→ α = -57.9 − (+24.5) = -82.4%
— what this call made or lost against just owning the index. This is the number the verdict uses.
Stock-selection read: -33.4% — did MSFT move the predicted way relative to the market. The two only differ on a SHORT: this call was right about direction relative to the index but still cost money versus holding it.
- Rule fired:
· HIT — R ≥ +10% AND α ≥ +5· PARTIAL — R ≥ +5% OR α ≥ 0▶ MISS — everything else
- Credit: supporters of a MISS earn 0 each, opposers the inverse — this feeds the scoreboard weights. supported: Chamath, Jason, Sacks | opposed: Friedberg
Conviction timeline
bands: green ≥ 65 · watch ≥ 45 · ember ≥ 15
Plays vs SPY · % since first mention (2023-02-17)
Plays
| expression | symbol | kind | relevance | rationale |
|---|---|---|---|---|
| ▼ SHORT | MSFT | stock | PRIMARY | OpenAI's listed partner and the operator of Bing/Sydney - the specific trust-and-safety layer (and the ChatGPT DAN jailbreak) under discussion; OpenAI is private |
| ▼ SHORT | GOOGL | stock | adjacent | Bard ships the same human-programmed filter layer and Google is the other half of the monoculture critique |
Mention log
Jason
politics-market w=1.00 · n=62
· E116 (2023-02-17) · sentiment · strength 3
▶ 1:19:52
SUPPORT
“And this filtering, and then it's giving different answers to different people about the same prompt. So this is what people are doing right now is trying to figure out, as you're saying, Sacks, what did they put into this? And who is making these decisions?”
Chamath
politics-market w=1.02 · n=96
· E116 (2023-02-17) · sentiment · strength 2
▶ 1:12:47
SUPPORT
“Well, to your point, I think that is the big existential risk with this entire part of computer science, which is why I think it's actually a very bad business decision for corporations to view this as a canonical expression of a product.”
Sacks
politics-market w=0.97 · n=88
· E129 (2023-05-19) · explicit_prediction · strength 3
▶ 42:28
SUPPORT
“So we're going to do the same thing here. We're going to buy into these phony narratives here to get the centers off our backs. And that's going to create this giant AI industrial complex that's going to slow down real innovation and be a burden on entrepreneurs.”
Who built this conviction
each voice's total force on the score — supports and opposes from every mention, weighted exactly as the replay applied them · share = % of all mention-driven movement
⏳ decay drained -61.5 over the idea's life — that's time passing, attributed to no one
Score events
| episode | kind | Δ | after | note |
|---|---|---|---|---|
| E116 2023-02-17 | init | +40.6 | 40.6 | E116 born by Sacks (explicit_prediction x3) [w=0.97] |
| E116 2023-02-17 | reinforce | +13.7 | 54.2 | E116 Chamath support x2 (new voice) [w=1.02] |
| E116 2023-02-17 | oppose | -12.1 | 42.2 | E116 Friedberg opposes x3 [w=0.84] |
| E116 2023-02-17 | reinforce | +15.6 | 57.8 | E116 Jason support x3 (new voice) |
| E118 2023-03-03 | decay | -2.3 | 55.5 | E118 silent |
| E119 2023-03-11 | decay | -2.2 | 53.2 | E119 silent |
| E120 2023-03-17 | decay | -2.1 | 51.1 | E120 silent |
| E121 2023-03-24 | decay | -2.0 | 49.1 | E121 silent |
| E122 2023-03-31 | decay | -2.0 | 47.1 | E122 silent |
| E123 2023-04-07 | decay | -1.9 | 45.2 | E123 silent |
| E124 2023-04-14 | reinforce | +6.3 | 51.6 | E124 Sacks support x1 [w=0.97] |
| E125 2023-04-21 | decay | -2.1 | 49.5 | E125 silent |
| E126 2023-04-28 | decay | -2.0 | 47.5 | E126 silent |
| E128 2023-05-12 | decay | -1.9 | 45.6 | E128 silent |
| E129 2023-05-19 | reinforce | +9.5 | 55.1 | E129 Sacks support x3 [w=0.97] |
| E130 2023-05-26 | decay | -2.2 | 52.9 | E130 silent |
| E131 2023-06-02 | decay | -2.1 | 50.8 | E131 silent |
| E132 2023-06-10 | decay | -2.0 | 48.7 | E132 silent |
| E133 2023-06-16 | decay | -1.9 | 46.8 | E133 silent |
| E134 2023-06-24 | decay | -1.9 | 44.9 | E134 silent |
| E135 2023-07-01 | decay | -1.8 | 43.1 | E135 silent |
| E136 2023-07-09 | decay | -1.7 | 41.4 | E136 silent |
| E137 2023-07-14 | decay | -1.7 | 39.7 | E137 silent |
| E139 2023-07-27 | decay | -1.6 | 38.1 | E139 silent |
| E140 2023-08-04 | decay | -1.5 | 36.6 | E140 silent |
| E141 2023-08-11 | decay | -1.5 | 35.2 | E141 silent |
| E142 2023-08-18 | decay | -1.4 | 33.7 | E142 silent |
| E143 2023-08-25 | decay | -1.3 | 32.4 | E143 silent |
| E144 2023-09-01 | decay | -1.3 | 31.1 | E144 silent |
| E146 2023-09-22 | decay | -1.2 | 29.9 | E146 silent |
| E147 2023-09-29 | decay | -1.2 | 28.7 | E147 silent |
| E148 2023-10-07 | decay | -1.1 | 27.5 | E148 silent |
| E149 2023-10-13 | decay | -1.1 | 26.4 | E149 silent |
| E150 2023-10-20 | decay | -1.1 | 25.4 | E150 silent |
| E151 2023-10-27 | decay | -1.0 | 24.3 | E151 silent |
| E152 2023-11-03 | decay | -1.0 | 23.4 | E152 silent |
| E156 2023-12-08 | decay | -0.9 | 22.4 | E156 silent |
| E157 2023-12-16 | reinforce | +11.2 | 33.7 | E157 Sacks support x2 [w=0.97] |
| E158 2023-12-23 | decay | -1.3 | 32.3 | E158 silent |
| E159 2023-12-29 | decay | -1.3 | 31.0 | E159 silent |
| E160 2024-01-06 | decay | -1.2 | 29.8 | E160 silent |
| E161 2024-01-13 | decay | -1.2 | 28.6 | E161 silent |
| E162 2024-01-19 | decay | -1.1 | 27.5 | E162 silent |
| E163 2024-01-26 | decay | -1.1 | 26.4 | E163 silent |
| E164 2024-02-02 | decay | -1.1 | 25.3 | E164 silent |
| E165 2024-02-09 | decay | -1.0 | 24.3 | E165 silent |
| E166 2024-02-16 | decay | -1.0 | 23.3 | E166 silent |