Sam Altman, of Y Combinator, said that, in the small world of Silicon Valley, the very idea of talent agents presents a “negative-selection problem”: “The actual 10x engineers don’t need or want an agent; people quickly discover they’re great, and they end up picking where (and especially with whom) they want to work. In my limited experience, the engineers that get agents are bad.”
That may be true now, but programming has a huge reputation management component. Negotiation isn't hard to learn, but it's always better to have an advocate than to negotiate for yourself (note: lawyers retain other lawyers) and detecting trends in technologies and finding work that helps a programmer's career is a full-time job in its own right.
By many measures, the star system didn’t work that well for Hollywood: it made moviemaking more expensive, which made studios more risk-averse, which led to inferior creative projects. And programmers are not movie stars—not yet, anyway. “Movie stars have their own brands,” McKinnon, the Okta C.E.O., said. “People will go to see a movie just because it has Tom Cruise in it. But programmers don’t really have that. No one’s going to pay for a product just because James Gosling built it.” (Gosling is one of the inventors of Java.) “Well, geeks like me will. But most people won’t. They pay for a service.”
In Hollywood and software both, value delivered is multiplicative rather than additive. A good actor in the lead role might add 10% to revenue. If it's going to be a $5 million film, that's only $500,000. If it's going to be a $250 million film, that's huge: now we're talking about $25 million. Of course, many of these multiplicative variables are latent and unknown until the thing's actually released.
Because of this, no one knows what "the base" is, or who contributed what, and there's wide room for negotiation.
That may be true now, but programming has a huge reputation management component. Negotiation isn't hard to learn, but it's always better to have an advocate than to negotiate for yourself (note: lawyers retain other lawyers) and detecting trends in technologies and finding work that helps a programmer's career is a full-time job in its own right.
By many measures, the star system didn’t work that well for Hollywood: it made moviemaking more expensive, which made studios more risk-averse, which led to inferior creative projects. And programmers are not movie stars—not yet, anyway. “Movie stars have their own brands,” McKinnon, the Okta C.E.O., said. “People will go to see a movie just because it has Tom Cruise in it. But programmers don’t really have that. No one’s going to pay for a product just because James Gosling built it.” (Gosling is one of the inventors of Java.) “Well, geeks like me will. But most people won’t. They pay for a service.”
In Hollywood and software both, value delivered is multiplicative rather than additive. A good actor in the lead role might add 10% to revenue. If it's going to be a $5 million film, that's only $500,000. If it's going to be a $250 million film, that's huge: now we're talking about $25 million. Of course, many of these multiplicative variables are latent and unknown until the thing's actually released.
Because of this, no one knows what "the base" is, or who contributed what, and there's wide room for negotiation.