Exploring the frontier // Nomad since '19 // Prev CMO @irys_xyz & @fermah_xyz
Joined July 2017
- Tweets 3,706
- Following 1,705
- Followers 7,236
- Likes 9,782
188 Photos and videos
Anthropic just pulled the oldest trick in SaaS pricing.
I pay $200/mo for Claude Max. My limits have been noticeably worse this past week.
Now they announce 2x off-peak usage for two weeks. Sounds generous.
But here’s what actually happens: limits quietly drop, a temporary 2x makes the reduced limit feel normal, the promo ends, and you’re left at a baseline lower than where you started. You just didn’t notice the downgrade because the 2x absorbed the transition.
These AI plans are massively subsidized. The raw compute behind a heavy user costs multiples of the subscription price. Every move like this is the subsidy quietly correcting.
Very sneaky, Anthropic.
A small thank you to everyone using Claude: We’re doubling usage outside our peak hours for the next two weeks.
415
369
8,035
1,469,846
Goddamn government, what did my 0 revenue producing vibecoded app ever do to you? 😭
Jun 13
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
1
167
The $200 AI plans are subsidized far harder than anyone assumed
SemiAnalysis bought the top Anthropic and OpenAI subscriptions and ran coding tasks until the weekly limits hit. At API prices, Claude's plan paid out roughly $8k in tokens. OpenAI's paid out roughly $14k.
That's a 40-70x subsidy on power users.
And it's still the cheapest way to acquire customers in software. Let builders overuse your model until their habits, repos, and prompts live inside it.
The labs are buying the daily-driver slot now. Rent will be due soon enough
Jun 10
Recently, we purchased one of each Anthropic/OpenAI subscription plan and randomly ran long horizon coding tasks until we exhausted the weekly limit. It's widely believed that a $200/month plan maxes out at ~$2000/month worth of tokens (assuming API pricing). However, we found that the subscriptions are actually far more generous. (2/4)
1
591
It's tedious, but take 30 mins to take money out of DeFi today
- For Solana: use Jupiter Portfolio to quickly find your DeFi positions
- For Ethereum: use Rabby to find your positions. Also, bulk revoke approvals from the wallet
DeFi is in for a rough few days and weeks if Mythos ends up being as deadly as advertised
And even if it's not, the potential reward for the risk you're taking on is horrendous
1
1
120
How about building vibe coded apps that generate zero revenue while backpacking across Asia?
Jun 8
funemployment used to mean backpacking across Asia for six months. now it means building 5 vibe coded apps that generate zero revenue
3
11
6,935
Just used @cursor_ai after a few months, and damn, it's gotten a lot better
Composer 2.5 is also a pretty decent model
111
The two weeks of unlimited Codex usage were glorious while they lasted
Jun 4
We're fixing a codex bug today that was causing us to undercount tokens being served to some Pro and Plus accounts by a small amount. This impacted < 15% of accounts.
Not the kind of bug you want us to fix, but didn't want to do this silently and thought you should know.
615
Ethnicity being the defining characteristic of someone’s background is way fucking weirder tbh
2
271
Powered by xAI’s compute
I’d bet that the real reason for not releasing Mythos was lack of compute, rather than ‘fear of model’s cyber capabilities’
May 28
JUST IN: Anthropic announces it will roll out Claude Mythos “in the coming weeks” despite growing fears over the model’s cyber capabilities.
6
327