Hacker News Reader

Comment by jstx1 | Hacker News Reader

Lex was the most obvious example of a bad interviewer that came to mind after I read the title of this post. He's had an impressive selection of people on his podcast but his communication and interviewing style is the worst I've ever come across - very dry delivery, not following up on what the guests says, pointless questions about the meaning of life or good and evil or scifi-technology that doesn't exist, reading from a list of prepared questions, trying too hard to push his own opinions at some points, sucking up too much to Joe Rogan and Elon Musk etc.

I really don't know how he managed to get the selection of guests that he's had especially in the early days of the podcast.

View on HN

>I really don't know how he managed to get the selection of guests

For the later guests who were not colleagues at MIT or his previous contract work at Google, they do what many guests do: they look at the interviewer's previous episodes and see if they like his style. It's a big commitment to spend ~3 hours with someone they don't know so they use the previous episodes as a "résumé" to see if it's worth their time.

>, not following up on what the guests says

And the parent comment you replied to said the opposite: >Lex will often say something like "oh that's interesting, let's talk about that"

Not debating you but it's an interesting observation to me that different people will see contradictory evidence.

>, reading from a list of prepared questions

The way a lot of these podcasts' preparations work behind-the-scenes with emails/phone calls before the actual interview is that the interviewer will propose a list of topics to talk about. The guest can also add (or subtract) from the list of topics because they are the ones who ultimately agree whether to appear on the podcast at all.

Therefore, using prepared questions is actually respecting the guest's agreement to be interviewed.

Now, if on the other hand you're complaining about the delivery of the prepared questions... such as "Lex should memorize the 50 prepared questions perfectly instead of reading a sheet of paper in front of him" or "Lex should use a cue card or teleprompter off camera to read the prepared question" ... it seems like a really trivial thing to complain about.

> they do what many guests do: they look at the interviewer's previous episodes and see if they like his style.

Unless you have some inside information, you can't know that this is the case. I find it difficult to believe given that some of his early guests are

Steven Pinker - episode 3

Guido van Rossum - episode 6

Eric Schmidt - episode 8

Elon Musk - episode 18

It's naive to think that you get a retired-Google-CEO-billionare by cold calling and getting him to review and like the first 7 episodes of your newly started podcast.

>It's naive to think that you get a retired-Google-CEO-billionare by cold calling

Hey, you cut out the quote where I conditionally qualified my statement with "later guests who were not colleagues at MIT or his previous contract work at Google,".

Anyway, it seems reasonable to presume that his connections at Google went up the chain and reached Eric Schmidt. So it wouldn't have been a true cold call.

He said Elon Musk happened because his A.I. research on Tesla self-driving was already seen by Elon.

Some of the early guests who were professors via his MIT connections. Some guests (especially physicists) might have been connections with his dad who is a physics professor. His older brother runs a biotech company so that may also spur some introductions.

With "6-degrees-of-separation" being what it is, what looks like a cold-call to outsiders ("how did he get Eric Schmidt?!?") could have actually been a friend-of-a-friend or colleague-of-a-colleague. I'm a nobody and I happen to know the billionaire fund manager and owner of a major sports team. With 3-degrees of separation with various past co-workers, I can connect the dots to several CEOs of major corporations. That doesn't mean they'd agree to be on my hypothetical podcast if I asked them but if a few of them said "yes", some viewers would wonder, "how did he manage to get X with a cold call?" -- The answer is that my extended social graph is invisible to viewers and it really wasn't a cold call.

Examples of "later" guests that are more "cold calls" with no help from MIT/Google were Dan Carlin (Hardcore History) and some of the female guests who weren't scientists. This isn't speculation because the guests themselves said they looked at Lex's previous podcasts to see his work.

In general, the playbook for a new podcaster (with no audience of millions yet) to get guests follows a similar trajectory:

1) Start with personal connections to build a body of work. Cold-calling famous people won't work at this early stage because they don't have any prior work to see to if it's worth their time.

2) The previous podcasts episodes can start to act as a résumé so host doesn't have to rely on personal connections. The ever-increasing audience numbers also pads the résumé and adds credibility to the interviewer.

3) The transition eventually happens where interviewer can get new guests based on reputation more than personal connections.

>Not debating you but it's an interesting observation to me that different people will see contradictory evidence

So what do you think that means in a deep, pilosophical way-of-life way?

There’s a lot of nuance here.

Lex, if you read this, I think improving the questions you ask is the biggest thing that would contribute to improving your interviews.

"The meaning of life" is an ok question, but asking every interviewee about aliens and sci-fi technology, regardless of background, is getting repetitive. It doesn't seem to really give you, or the audience much to work with, and basically encourages the interviewee to spout off random stuff.

Now, the Jim Keller interview, that was good stuff because Jim Keller is an interesting person and a good people manager.

I have no idea whether this applies to Lex, or if there are better explanations, but... he probably just asks. In my limited experience with my little podcast in the sports analytics world, I'm consistently surprised by who says yes, and how few people say no. ~90% of the people I've asked to be a guest have said yes, and those who have said no haven't done so because they're "too famous" or something like that.

I think that he's had way too many billionaires, top researchers, authors, computer scientists, and celebrities to attribute it to just asking.

Working at MIT and having a small history of good guests probably helps.

The guests themselves make value judgments of whether this polite, motivated, knowledgeable, reasonably boring guy with a sizeable audience and good credentials can help them to spread their own messages without much journalistic pushback.

His audio quality is also not bad.

-----

Personal opinion: The interviews are far too long. I'd like an informed opinion of the best (as opposed to most famous or controversial or odd) guests. If a biology expert told me Lex's interview with Eric Lander was a Must Listen, I'd value that more than, "oh! It's three hours of Elon Musk, but his personal values are well-known and haven't changed."

I would listen to a few very interesting conversations unedited OR a larger collection of reasonably interesting, short, heavily curated conversations. I enjoy hearing about subjects I normally don't encounter, and some of his guests bring this experience. Listening to months of audio to find these gems is not for me.

I prefer a small group of podcasts that have shorter episodes, entertaining/engaging hosts, interesting guests,

Personally I find 20 to 30 minute interviews the sweet spot. Sometimes I’ll go a bit shorter and sometimes a bit longer but for 1:1 something like an hour feels too much most of the time.

I like the long conversations, because it’s easier to be a fake person for 30 minutes, but comparatively it’s hard to fake being a genuine person or expert on a topic for 2 or 3 hours.

I'm sure it's a fault of my attention span but I tend to be a fan of 30 minute conference presentations and otherwise distilling topics down to about that length. Otherwise, it often feels like you're dragging things out. I'm not having a date with the person. I want to discover for my audience what key insights they have.

I do sometimes have longer podcast conversations with people I know well but 20-30 minutes feels like the natural sweet spot for my typical interview.

I want brisket that is delicious after you microwave it, but it tends to require a long time in the oven.

But I don't have 2 or 3 hours to listen in to conversations of strangers and then decide if they are worthwhile or not.

I was also drawn in by Lex's selection of guests, but haven't been listening since his interview with Yeonmi Park. I was attentive for the first half, but she abrubtly refers to LGBT culture as regressive and the conversion devolves into rambling praise for Jorden P., who happens to have also had her on his podcast recently (see the comment on Frequency Bias above).

Well yeah he isn’t the most polished, but he puts the guests at ease and the conversations benefit from that. He reminds me of Tim Ferris in some ways, who was often barely coherent, although Ferris was never quite believably authentic and was clearly building some kind of empire.

Reply on news.ycombinator.com