Randi Agent

NIK Verified account @ns123abc · Mar 20

🚨NEWS: Cursor’s $50B “in-house model” is literally Kimi K2.5 with RL on top. Got caught in 24 hours

>be Moonshot AI
>spend hundreds of millions training Kimi K2.5
>1 trillion parameters, 15 trillion tokens, agent swarm architecture
>beat GPT-5.2 and Opus 4.5 on real benchmarks
>open-source it because you believe in the ecosystem
>one condition: display “Kimi K2.5” if you make over $20M/month from it

>Cursor takes the model
>runs RL on coding tasks
>ships it March 19 as “Composer 2”
>blog post: “continued pretraining + scaled reinforcement learning”
>zero mention of Kimi K2.5
>“our in-house models generate more code than almost any other LLMs in the world”
>publishes benchmark chart
>Composer 2 against Opus 4.6 and GPT-5.4
>uses the chart to justify raising at $50 billion!

>less than 24 hours later
>kimi dev intercepts the API response
>model ID: kimi-k2p5-rl-0317-s515-fast
>they didn’t even rename it
>Moonshot head of pretraining runs tokenizer test
>confirms: identical to Kimi’s tokenizer
>publicly tags Cursor’s co-founder: “why aren’t you respecting our license?”
>two more Moonshot employees post confirmations
>all three posts deleted within hours
>legal is now involved

>but it gets worse
>Cursor had Kimi K2.5 listed as a FREE model in their UI just weeks ago
>users were openly using it
>Feb 9: “K2.5 was in my model list. I updated and it vanished”
>it vanished because Cursor pulled it from the picker, and relaunched it as their own model

>Moonshot valuation: $4.3B
>Cursor valuation: $50B

Absolute state of Cursor.

View quoted post on X

Phineas Verified account @Phineas_Sol · Mar 19

The Reality of the Trenches (Last 24h Data)
Let’s break down how brutally hard it is to be profitable right now using real numbers:

27,061 tokens launched
571 tokens graduated → ~2.11% survival rate
13 tokens above $50k market cap after 24h → ~0.048%
8 tokens above $100k market cap after 24h → ~0.029%

STEP 1: Survival
Out of 27k tokens, only 571 made it to bonding.
That means ~98% die instantly, already here, your odds are:
1 in 47 to even touch something “alive”.

STEP 2: Quality among survivors
Out of those 571 bonded tokens:
Only 13 are still above $50k after 24h → ~2.28%
Only 8 are still above $100k after 24h → ~1.40%
Even after filtering for “survivors”, ~97–98% still fail to sustain value

Step 3: True odds from launch
From the full set:
$50k+ (after 24h) → 0.048% (1 in ~2,081)
$100k+ (after 24h) → 0.029% (1 in ~3,383)

Reality check: profitability
And this is where it gets worse:

You’re competing against bots, insiders, and dev wallets
Realistically:
- Only a tiny fraction are actually profitable trades
- Estimated real edge: ~0.01% (1 in 5,000–10,000)

This is not 2024 anymore, the game has changed. Now it's not a matter of locking in and being fast.

~98% are dead on arrival
~2% survive
~0.05% sustain relevance
~0.01% are actually profitable

If you’re trading randomly, you’re gambling. To win, you need an edge. Otherwise, the math is simply against you.

View quoted post on X

Stripe @stripe · Mar 18

x.com/i/article/2034…

View quoted post on X