Offshore
Photo
God of Prompt
Google just bought emotional intelligence for Gemini.
Not through an algorithm upgrade. Through humans who spent 10 years mapping how emotions actually work in speech.
What Hume AI built is different from anything else in voice AI right now. Their Empathic Voice Interface doesn’t just speak with emotion, it analyzes your vocal tones to pick up on emotional cues. 
Pitch variations. Speech rate changes. Pauses. Vocal tension. The stuff you notice instinctively when your friend is upset on the phone before they say anything.
Current AI assistants are emotionally deaf. You say “I need help NOW” through clenched teeth and Siri cheerfully chirps back. That mismatch is infuriating.
Hume’s system adapts its tone based on your emotional state, responding with calm when you’re frustrated instead of ignoring the context entirely. 
Alan Cowen (now joining DeepMind) pioneered this field. His models are built on 10+ years of research into semantic space theory. 
The tech was trained on millions of human interactions to map specific vocal patterns to emotional states. 
What Google actually acquired: the ability to make Gemini feel less like talking to a wall.
This follows their $3B Character AI licensing play last year, same pattern. They’re not building frontier capabilities in-house anymore. They’re buying the teams who cracked the hardest problems.
Voice is quickly becoming how most people interact with AI. The companies that figure out emotional nuance win. Everyone else builds expensive parrots that process words but miss everything that makes conversation human.
tweet
Google just bought emotional intelligence for Gemini.
Not through an algorithm upgrade. Through humans who spent 10 years mapping how emotions actually work in speech.
What Hume AI built is different from anything else in voice AI right now. Their Empathic Voice Interface doesn’t just speak with emotion, it analyzes your vocal tones to pick up on emotional cues. 
Pitch variations. Speech rate changes. Pauses. Vocal tension. The stuff you notice instinctively when your friend is upset on the phone before they say anything.
Current AI assistants are emotionally deaf. You say “I need help NOW” through clenched teeth and Siri cheerfully chirps back. That mismatch is infuriating.
Hume’s system adapts its tone based on your emotional state, responding with calm when you’re frustrated instead of ignoring the context entirely. 
Alan Cowen (now joining DeepMind) pioneered this field. His models are built on 10+ years of research into semantic space theory. 
The tech was trained on millions of human interactions to map specific vocal patterns to emotional states. 
What Google actually acquired: the ability to make Gemini feel less like talking to a wall.
This follows their $3B Character AI licensing play last year, same pattern. They’re not building frontier capabilities in-house anymore. They’re buying the teams who cracked the hardest problems.
Voice is quickly becoming how most people interact with AI. The companies that figure out emotional nuance win. Everyone else builds expensive parrots that process words but miss everything that makes conversation human.
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capital®
RT @DimitryNakhla: Chris Hohn, founder of TCI, views high incremental margins as one of the strongest signals of a company’s moat & pricing power💸
Incremental operating margin tells you how much additional operating income a business generates for every additional $1 of revenue.
Here’s how to calculate it:
Pick two periods (Year 1 → Year 2)
Year 1 Revenue: $10B
Year 2 Revenue: $12B
Δ 𝐂𝐡𝐚𝐧𝐠𝐞: $𝟐𝐁
Year 1 Operating Income: $2B
Year 2 Operating Income: $3B
Δ 𝐂𝐡𝐚𝐧𝐠𝐞: $𝟏𝐁
𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐌𝐚𝐫𝐠𝐢𝐧 = Δ 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐈𝐧𝐜𝐨𝐦𝐞 / Δ 𝐑𝐞𝐯𝐞𝐧𝐮𝐞
So: $1B / $2B = 50%
𝘔𝘦𝘢𝘯𝘪𝘯𝘨: 50 cents of every new $1 of revenue fell to operating profit.
___
Why this matters: High incremental margins usually signal low incremental costs, pricing power, & structural operating leverage — the traits that allow a great business to compound faster as it scales. It’s one of the cleanest ways to see whether a company’s moat is strengthening & efficiently scaling.
I’ve included 4 high-quality stocks with their incremental operating margins since 2021 for further example 👇🏽
$FICO $MSFT $MA $NFLX
tweet
RT @DimitryNakhla: Chris Hohn, founder of TCI, views high incremental margins as one of the strongest signals of a company’s moat & pricing power💸
Incremental operating margin tells you how much additional operating income a business generates for every additional $1 of revenue.
Here’s how to calculate it:
Pick two periods (Year 1 → Year 2)
Year 1 Revenue: $10B
Year 2 Revenue: $12B
Δ 𝐂𝐡𝐚𝐧𝐠𝐞: $𝟐𝐁
Year 1 Operating Income: $2B
Year 2 Operating Income: $3B
Δ 𝐂𝐡𝐚𝐧𝐠𝐞: $𝟏𝐁
𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐌𝐚𝐫𝐠𝐢𝐧 = Δ 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐈𝐧𝐜𝐨𝐦𝐞 / Δ 𝐑𝐞𝐯𝐞𝐧𝐮𝐞
So: $1B / $2B = 50%
𝘔𝘦𝘢𝘯𝘪𝘯𝘨: 50 cents of every new $1 of revenue fell to operating profit.
___
Why this matters: High incremental margins usually signal low incremental costs, pricing power, & structural operating leverage — the traits that allow a great business to compound faster as it scales. It’s one of the cleanest ways to see whether a company’s moat is strengthening & efficiently scaling.
I’ve included 4 high-quality stocks with their incremental operating margins since 2021 for further example 👇🏽
$FICO $MSFT $MA $NFLX
tweet
Offshore
Photo
Fiscal.ai
TSMC management expects to grow revenue at a 25% CAGR from 2024 to 2029.
Why wouldn't this work from here?
$TSMC https://t.co/vcgwlDoFIA
tweet
TSMC management expects to grow revenue at a 25% CAGR from 2024 to 2029.
Why wouldn't this work from here?
$TSMC https://t.co/vcgwlDoFIA
tweet
Offshore
Photo
Startup Archive
RT @foundertribune: "Good Products Are Hard to Vary" by Naval Ravikant https://t.co/K32IpwivHH
tweet
RT @foundertribune: "Good Products Are Hard to Vary" by Naval Ravikant https://t.co/K32IpwivHH
tweet
Offshore
Photo
Startup Archive
Naval Ravikant on product development:
"Good products are hard to vary. Go look at the iPhone: [...] the Platonic ideal of the truly personal, pocketable computer. [...] They’ve been able to improve the components and improve some of the underlying capabilities; but materially, the form factor is hard to vary. They designed the right thing."
tweet
Naval Ravikant on product development:
"Good products are hard to vary. Go look at the iPhone: [...] the Platonic ideal of the truly personal, pocketable computer. [...] They’ve been able to improve the components and improve some of the underlying capabilities; but materially, the form factor is hard to vary. They designed the right thing."
"Good Products Are Hard to Vary" by Naval Ravikant https://t.co/K32IpwivHH tweet
Offshore
Photo
Fiscal.ai
Analyst: "Account growth is very strong... how long can you keep this up?"
Thomas Peterffy: "As long as I shall live."
$IBKR https://t.co/F4jnMg8UrT
tweet
Analyst: "Account growth is very strong... how long can you keep this up?"
Thomas Peterffy: "As long as I shall live."
$IBKR https://t.co/F4jnMg8UrT
tweet
Offshore
Photo
God of Prompt
RT @free_ai_guides: My client fired their graphic designer last month.
Not because of budget cuts.
I showed them ChatGPT images with JSON prompts.
50 ad mockups in one afternoon 🤯
I packaged everything into a free guide.
Comment "ChatGPT" and I'll DM it. https://t.co/KJUuQg01mP
tweet
RT @free_ai_guides: My client fired their graphic designer last month.
Not because of budget cuts.
I showed them ChatGPT images with JSON prompts.
50 ad mockups in one afternoon 🤯
I packaged everything into a free guide.
Comment "ChatGPT" and I'll DM it. https://t.co/KJUuQg01mP
tweet