What questions can tech CEOs be asked, to guage their moral trustworthiness?
In light of the OpenAI shakeup, many here have pointed to Altman's goals with the project being ones most don't consider the most important (money, reckless use which many enhance government/powerful control), and a 'personality that is required for this kind of thing'.
What questions or things can we talk about with tech CEOs (and others in decision-making positions, to get a grasp of their morality/personality?
69 comments
[ 3.0 ms ] story [ 159 ms ] threadIf you haven't had a lot of practice, ask people who have. Investors and other CEOs typically have lots of practice, though sometimes they aren't trustworthy themselves. So ask several people and discard outlying opinions.
Ignore the press, unless there are external facts like a conviction. The press is full of snark and innuendo about CEOs, most of which is just reporters trying to cook up a narrative people will click on.
If they say anything positive, you know to run the other direction.
Does that include religious people who don't have view their creator as malevolent?
I suspect for this particular heuristic to feel sensible, the god needs to be one you happen to believe in. I would take precisely zero comfort in knowing that someone fears an invisible, likely bronze age supernatural authority with dozens of permutations of subjectively interpreted doctrine and morality.
But I could be overthinking this.
I think, WHICH moral code you choose has some relevance here.
How can this be done, and how can you tell when they're in that "state"?
But in the case of interviewing a CEO, you don't often have a chance to see them interacting with random people.
People can present themselves differently than they are, and especially people with lower morales will have no problem presenting themselves differently if that is in their interest. Look at pretty much any big company CEO.
If you do need to gauge someone, look at their history, look at their actions, and look at their incentives. If they don't align with what they're saying you should not expect them to act any different than they did before.
This applies not only to CEOs but to every single human on the planet. One thing I have learned is that many people seem to be easily tricked by certain types of humans saying one thing and doing another (which is contradictory to the "said" thing). It is certainly frustrating to watch happen, when it involves people you care about.
Seems to me that almost everybody. And the ones not tricked just happened to be looking at a different place by chance, and won't replicate the same result a different time.
We definitively need some kind of tooling to deal with this.
- "Should Palestine should be free?"
- "Do Black Lives Matter?"
That being said, asking a CEO to take responsibility for a mistake is a pretty good indicator.
If they own the mistake. That's good. If they own the mistake and try to fix it. That's better. If they own the mistake and ask the people wronged how to make amends. They're probably not a CEO.
You can't just say the magic incantation "I take full responsibility" and leave it at that.
A decent rule of thumb is that if someone has to overtly say "I take full responsibility", they aren't taking full responsibility.
You really have to take off your rose-tinted glasses and try to get to know what their core values are. That is the only way.
Tell them that their response is going to be publicized beforehand.
My answer would be to check their technical background to see how they ended up as CEO. Was it circumstance, or raw machiavellian ambition?
If you must ask one question I would probably ask, "Do you hate being a CEO?". If they answer "Yes" it likely means they are doing it out of responsibility rather than pathalogical persuit of the unattainable.
1. https://en.wikipedia.org/wiki/Dodge_v._Ford_Motor_Co.
What are your views on unionization? Would you voluntarily recognize an union formed by your employees?
Why don't our job postings have salary ranges?
Remote work?
Open-source?
Climate change?
How do you think our new office will impact the local community and housing affordability?
What philanthropic activities do you engage in? What about our company?
How will our culture change once our company achieves success?
What can we get in writing, in revised bylaws, rather than a verbal promise?
This kind of approach would mean you also look at governance structures, which in the case of OpenAI would have focused on the issues with their board structure.
For most people who get the chance to talk with "tech CEOs", the basic context of the conversation is usually built on a profit-seeking foundation -- Should I invest in you? Should I sign a commercial deal with you? Should I work with you as an employee? On that basis, morality is often orthogonal to profit (sometimes opposed, and rarely aligned with).
Most people, most of the time, if they are evaluating tech CEOs for any semblance of "morality" or ethics, it is usually most pertinent to consider business ethics -- Will you uphold the terms of a contract? Will you sincerely work to create shareholder value (above personal goals or benefits)? Will you be a good leader for your employees, treat them fairly according to professional standards and company policies? This level of ethics is necessary in order to conduct good business; trust is currency and if you cannot trust a CEO to uphold standard business and professional ethics, then the cost of doing business in an environment of mistrust usually becomes exceedingly high.
Anything beyond that, like is this CEO a "good person", are they working for the betterment of mankind, etc., is usually not worth evaluating for 95% of people -- not because it's unimportant, but because (a) capitalism is amoral and grafting morality on to it is kind of an exercise in futility, and (b) most people are not equipped to evaluate the answers in any kind of serious or logically rigorous way.
Auto-fail anyone who treats the driver poorly. Find the guy that wants and/or tries to help the driver succeed.
Various ways to see the inner person and not the "show person" on display for the interview.
But that's pretty funny to learn :) Nice.