Treat Startup Equity as Zero, Then Decide Whether to Go

When choosing a startup offer, first value the equity at zero.

X user GenAI_is_real gave a friend even harsher advice in a post: first ask yourself whether you can accept the equity going to zero, or even turning into debt. In extreme cases, startup equity can be worth less than zero. Paper wealth creates a feeling of wealth that is almost magical, but when evaluating an offer, valuing equity at zero is a risk calibration, not a sign of pessimism about startups.

The reasoning is not complicated. Many star startups have high turnover, and after employees leave, their shares usually cannot be sold until the company IPOs. Recently, many Silicon Valley companies have started doing tender offers, meaning the company buys back old shares, but the people who really benefit are mostly a small group who joined early and stayed. They endured the early psychological toll, and it is fair to say they earned the return. Most employees either did not last and left, or joined later. By the time a company starts doing tender offers, it is usually growing fast and has many new employees, who often do not qualify. Valuing equity at zero makes choosing an offer much calmer.

The poster added another layer: even after doing this extreme thought experiment, he still works at RadixArk and believes in the team and mission. He does not think RadixArk's equity is truly zero; the trust and shared responsibility he has built with the team make him believe in the value everyone is creating. Valuing equity at zero is a reminder for friends choosing startups, and also what he would say to himself if he went back to the day he chose his offer.

A more important factor is leadership that is worth following long term. Silicon Valley has many smart people and a lot of packaging, but few leaders are truly worth following. Some leaders' core teams follow them from big companies to startups, back to big companies, and then to another startup. Wherever the person goes, the core team follows, and the decision-making stays consistent. This kind of leadership does not rely on rhetoric. In career growth, having someone believe in you and lift you up is very practical. When you meet such a person, loyalty stops being a joke.

In the same timeline, Jinyan Su's 2026 job-search post offers another angle. She writes about her experience as an ordinary PhD looking for a job. In the comments, someone asks whether it is still possible to switch into the post-training/RL direction now, and says they do not know which area to study. Someone else asks how to relieve job-search anxiety. Jinyan Su's answer is not chicken soup, and is even a bit cruel: the only thing that can relieve it is finding a job; after she got an offer, she immediately went into vacation mode and only felt better after resting for a few weeks; without an offer, either give yourself a break—though that is very hard to do—or see a psychiatrist and get medication for depression or anxiety.

The comments also discussed connections. Someone wondered whether friends sometimes push internal referral roles at their company that have not been publicly posted—this also happens at American companies in China—and whether this is common in North America. Hidden roles circulate through circles, and information asymmetry itself is an advantage. Connections do not replace ability, but they change the order in which you see roles. Someone else reminded people that those doing AI4bio should also note that pharma companies basically do not provide work visa sponsorship.

Putting the two pieces of advice together, the order is clear: first rule out the worst financial outcome, then judge the people. Valuing equity at zero is the financial bottom line; whether the leadership is worth following is the long-term variable. As for direction, anxiety, and connections, they will not disappear because of one piece of advice, but at least they can be put on the table and discussed.

Original post: GenAI_is_real

Jinyan Su's job-search post: My 2026 job search experience as an ordinary PhD

Quoted tweet: SuJinyan6