If @grok bot had a publish / subscribe feature it would go viral.
Imagine you share a link to a bot thread that someone else’s @grok bot can subscribe to and it receives the messages any time your bot publishes something.
@elon who’s building this?
https://x.com/joemccann/status/2112218153124262140
Imagine you share a link to a bot thread that someone else’s @grok bot can subscribe to and it receives the messages any time your bot publishes something.
@elon who’s building this?
https://x.com/joemccann/status/2112218153124262140
❤1
55% of Americans Call AI Slowdown a Good Thing
13% Call It Bad
@Reuters & @Ipsos
https://x.com/joemccann/status/2112269328808907094
13% Call It Bad
@Reuters & @Ipsos
https://x.com/joemccann/status/2112269328808907094
👀2
The electron is the bottleneck.
@morganstanley and @Jefferies Both Find More Than Half of 2026-2028 GPU Servers May Have Nowhere to Plug In; 2 GW of 2026 Completions Called Impossible
Two independent research teams reached the same conclusion on the US data center power gap.
@morganstanley estimated the shortfall in available power and concluded that more than half of the GPU servers sold between 2026 and 2028 might not have anywhere to be plugged in.
Jefferies used satellite imagery from SynMax Intelligence to track US data center construction sites and found between 16 and 18 gigawatts of gross capacity that can be energized this year.
SynMax pushes out timelines when land isn't cleared on schedule regardless of company claims.
By its measures, around 2 GW of US data center projects scheduled for completion in 2026 haven't begun land clearances, making them "impossible" to deliver on time.
https://x.com/joemccann/status/2112269957115609481
@morganstanley and @Jefferies Both Find More Than Half of 2026-2028 GPU Servers May Have Nowhere to Plug In; 2 GW of 2026 Completions Called Impossible
Two independent research teams reached the same conclusion on the US data center power gap.
@morganstanley estimated the shortfall in available power and concluded that more than half of the GPU servers sold between 2026 and 2028 might not have anywhere to be plugged in.
Jefferies used satellite imagery from SynMax Intelligence to track US data center construction sites and found between 16 and 18 gigawatts of gross capacity that can be energized this year.
SynMax pushes out timelines when land isn't cleared on schedule regardless of company claims.
By its measures, around 2 GW of US data center projects scheduled for completion in 2026 haven't begun land clearances, making them "impossible" to deliver on time.
https://x.com/joemccann/status/2112269957115609481
🔥4
Small-Cap Positioning at 16th Percentile, Deepest Underweight in the Current Cycle
Small-caps are one of the few equity segments where positioning is genuinely clean to the short side.
That creates a mechanical setup where any positive catalyst (rate-cut signaling, improvement in credit conditions, or a risk-appetite surge) can trigger short-covering rather than incremental conviction longs.
The 16th percentile is not the most extreme short on record but it's deep enough to matter in a squeeze scenario.
https://x.com/joemccann/status/2112326441090379892
Small-caps are one of the few equity segments where positioning is genuinely clean to the short side.
That creates a mechanical setup where any positive catalyst (rate-cut signaling, improvement in credit conditions, or a risk-appetite surge) can trigger short-covering rather than incremental conviction longs.
The 16th percentile is not the most extreme short on record but it's deep enough to matter in a squeeze scenario.
https://x.com/joemccann/status/2112326441090379892
US Long/Short Net Leverage at 43.7%, a Five-Year Low; Fell 3.2pts Last Week, Largest Drop Since Liberation Day
@GoldmanSachs prime brokerage data shows US long/short net leverage fell 3.2 points last week, the largest weekly decline since Liberation Day and the fourth-largest over the past five years.
Essentially all of the net exposure reduction since end of July has come from an increase in shorts, not a reduction in longs.
Short exposure is now near five-year highs while long exposure remains in the 53rd percentile over the past three years and the 72nd percentile over five years.
That means long books haven't been meaningfully reduced yet. If performance pressure continues, there's scope for further de-grossing through long sales alongside short covering.
https://x.com/joemccann/status/2112349454661730790
@GoldmanSachs prime brokerage data shows US long/short net leverage fell 3.2 points last week, the largest weekly decline since Liberation Day and the fourth-largest over the past five years.
Essentially all of the net exposure reduction since end of July has come from an increase in shorts, not a reduction in longs.
Short exposure is now near five-year highs while long exposure remains in the 53rd percentile over the past three years and the 72nd percentile over five years.
That means long books haven't been meaningfully reduced yet. If performance pressure continues, there's scope for further de-grossing through long sales alongside short covering.
https://x.com/joemccann/status/2112349454661730790
❤2