ohxiyu
Feed
Feed
A person's state begins to change not because they suddenly encounter good luck, but because they start to actively change their course of action.
Many times in life, it's not a lack of opportunities, but staying too long in familiar environments, unable to access new information, new people, and new possibilities.
When you start to actively adjust:
Changing your work environment, meeting different people, entering new social circles, learning new skills, trying new lifestyles—all essentially increase the probability of encountering opportunities.
Many changes do not bring immediate results, but they quietly change your cognition, abilities, and range of choices. The people you interact with change, the world you see changes, and the decisions you make will change accordingly.
What is called "good luck" is often not fate suddenly favoring you, but you actively moving toward places where good things are more likely to happen.
Life will not improve automatically by waiting.
True turning points often happen after you are willing to leave your comfort zone, accept uncertainty, and keep taking action.
You step out, and only then does the world have a chance to respond to you.
The group that uses AI the most is not very willing to pay.
There is a post on V2EX that surveys programmers' monthly AI expenses.
Most of them use it for free, a few have company reimbursements, and those who pay out of pocket mainly use proxy services and subscriptions.
Programmers naturally have a whole set of money-saving methods—free quotas, proxy services, running open-source models themselves; they know all the loopholes others can't find.
Plus, since they already know how to code, AI saves them time, but the perceived value is diluted by the thought "I can write it myself."
Non-programmers are actually more willing to spend. For tasks like making PPTs, writing materials, and organizing spreadsheets, they have no alternatives, so the two hours saved are real, and when calculated hourly, they dare to pay.
There is a harsh number about A-share IPO subscriptions: on average, you only win once every five years.
Based on the online winning rate of 0.028% in the first half of 2026, there are about 40 new stocks in Shanghai and Shenzhen each year, with a maximum subscription of 10,000 shares per stock, expecting to be allocated 2.8 shares, totaling 112 shares per year, which equals 0.2 times winning per year.
Winning once is a real profit. The average first-day gain in the first half of the year was 280.7%, with zero break-even failures, earning tens of thousands of yuan each time. But 280% is the ten-year peak since 2017, and the break-even failure rate exceeded 20% in 2022 — the one you get in the next five years probably won’t look like this.
So the actual form of maximum subscription for new shares is: holding a stock exposure of 170,000 yuan, exchanging it for a lottery ticket with an annualized expectation of about 10,000 yuan, but most likely no returns for several consecutive years.
Whether to bet or not depends on whether you can endure having nothing for those middle years.
Luke has been against ordinals for three years, and the direction was wrong from the start.
His stance hasn't changed in three years: inscriptions are abusing block space, filters on Knots. Some nodes do follow this, and Knots' share has increased quite a bit in the past two years, but the hash power side has basically been inactive.
The problem is, what he wants to deny is essentially the same as Bitcoin's own rules.
You pay a fee, miners include it in the block, that's the only admission threshold.
Once you start dividing "legitimate transactions" and "junk data" by purpose, small payments get hit first—why should a few-dollar transfer occupying the same block space be considered necessary?
Miners' accounts are more straightforward.
In 2023, F2Pool tried filtering but withdrew after a few days;
MARA simply launched Slipstream to exclusively accept non-standard transactions.
Fee income is becoming increasingly important, so it's impossible to get them to voluntarily charge less.
In the end, the filter doesn't block inscriptions, it blocks Luke's own influence.
As Coldcard co-founder deletes posts, over $130 million in Bitcoin stolen
Rodolfo Novak (NVK), co-founder of Coldcard manufacturer Coinkite, is deleting posts on his X platform and may remove website content; previously, a vulnerability in the Coldcard hardware wallet that existed for five years allowed hackers to steal over $130 million in Bitcoin.
Bitcoin developer Peter Todd and others have noticed these deletions, and volunteers are working to preserve evidence.
This is a major security incident affecting hardware wallet users, exposing a fundamental flaw in the private key generation process that has existed for years.
The deletions have raised questions about the company's transparency in handling the issue, potentially damaging trust in the Bitcoin ecosystem and hardware wallet security.
The vulnerability stems from insufficient entropy during private key generation over more than five years, allowing attackers to brute-force insecure private keys.
NVK has not completely cleared his personal timeline but selectively deleted posts, keeping his pinned apology; Coinkite denies improper removal of disclosed content from the website.
