TickerTain
TickerTain
NewsroomShortsPortfolioConvergence
NewsroomShortsPortfolioConvergence
←
▶ 1:41 · $MOONSHOT-AI · U.S. AI Execs Warn as China’s AI Push Grows, Substack Adds AI Detection Tool | Diet TBPN
episode briefing
TBPN

U.S. AI Execs Warn as China’s AI Push Grows, Substack Adds AI Detection Tool | Diet TBPN

2026-07-21 · 9 company · 10 thematic
sentiment
4 bull0 bear5 neu
speakers
john coogan

Co-host of TBPN, a live technology and business show focused on startups, venture capital and Silicon Valley.

quote
“The emergence of highly capable open autonomous AI systems including Moonshot AI's Kimmy K3 model and Alibaba's Quen 3.8 Max... Kimmy K3 was also competitive with US models on some benchmarks.”
—john coogan
now playing · $MOONSHOT-AI
$MOONSHOT-AIbullish· mediumjohn coogan
Moonshot's Kimi K3 model competitive with top US models on benchmarks
Chinese startup Moonshot AI's new Kimi K3 model matches US frontier models on benchmarks, backed by Alibaba's deep pockets and infrastructure.
$BABA···bullish· mediumjohn coogan
Alibaba positioned as China's Google in AI with massive Moonshot investment
Alibaba is the dominant cloud/AI player in China akin to Google, with a large stake in Moonshot AI and its own competitive Qwen model family.
$GOOGL···neutral· mediumjohn coogan
Google Flash Cyber aims to undercut Anthropic's Mythos GPT Cyber on cost for enterprise security
Google is targeting the high-cost AI cybersecurity niche with a cheaper fine-tuned Flash model, challenging Anthropic's expensive but benchmark-leading Mythos GPT Cyber.
$NVDA···bullish· mediumjohn coogan
Nvidia's Nemotron 3 Ultra gaining traction as Jensen invests up and down the model stack
Nvidia is building its own competitive model family (Nemotron) and investing across the AI stack, creating a new value capture layer beyond just chips.
$DEEPSEEKneutral· mediumjohn coogan
DeepSeek serves as base model for rapid fine-tunes by Perplexity and others
Chinese model DeepSeek is being used as a foundation for quick fine-tunes (e.g., Perplexity's '1776' model), raising questions about distillation enforcement and supply chain attr…
$PERPLEXITYneutral· lowjohn coogan
Perplexity rapidly fine-tuned DeepSeek into '1776' model demonstrating open-weight agility
Perplexity showcased the speed of building on open-weight Chinese models by releasing a fine-tuned DeepSeek variant quickly.
$SUBSTACKneutral· lowjohn coogan
Substack launches AI detection via Pangram integration to verify human-authored content
Substack integrates Pangram's AI detection to scan posts and comments, betting that disclosed human authorship becomes a premium signal in the creator economy.
$PANGRAMneutral· lowjohn coogan
Pangram powers Substack's new AI content detection feature
Pangram's detection technology is being productized through Substack to surface AI-generated content across the platform.
$RAMPbullish· mediumjohn coogan
Ramp's model router called 'god tier product expansion' addressing CFO token cost pain
Ramp's new model router automatically selects optimal models for tasks, solving enterprise frustration with token costs and model proliferation — potentially a new core product li…