Brendan Eich on the Lex Fridman Podcast [video]
youtube.com
youtube.com
I think Brendan Eich has got to be one of my favourite people to hear anecdotes about early web history. He was around at some pivotal moments, was in with the business side of things but also super technical. He’s also just very good at talking, he never appears to talk himself into a corner, somehow always avoids getting into the weeds, there’s a lucidity to all his answers that’s somehow very impressive. I wonder how much preparation he puts into his interviews.
As for Lex, personally I really like his style. He operates at a different pace to many of the people he’s interviewing, but I find this often enables him to see interesting parallels and analysis. It’s not always perfect, and maybe there’s room for improvement, though it gives the whole show an authenticity that it might not otherwise have if it were too polished.
I feel like his style works very well with all the guests he has on. They all seem to get on very well, which has a massive positive effect on how the guest interacts. Sure he has some prepared questions, but he also goes off piste quite a bit, and when he gets back to the prepared questions, I generally find they are well researched and very relevant.
I’d also like to know how he gets so many great guests on the show.
My initial reaction was that, of the people you listed, Steve Wosniak would be the best fit. I attended a lecture he gave a number of years ago, it was very inspiring. He’s not very ‘corporate’ so I think he’d get on really well with Lex.
But I am perhaps showing my bias here, I’d also be very interested to see how versatile Lex’s interview style is with Gates, Bezos and Zuckerberg.
So far I’ve only listened to a few. I mostly listen via the podcast feed.
I posted the ones I liked to my linkblog (https://links.markjgsmith.com):
[2020-12-26] Lex Friedman Podcast #80 - Vitalik Buterin - Ethereum, Cryptocurrency, and the Future of Money - Fascinating and well paced conversation with the founder of Etherium covering a wide range of topics including Satoshi Nakamoto, blockchains, proof of work and identity, PKI and digital signatures, Bitcoin, money, the Etherium origin story, smart contracts, software engineering and project governance challenges, proof of stake and consensus algorithms, sharding of storage and computation, Etherium 2.0, games built using smart contracts, Uniswap, AI and crypto, and closes on Immortality
https://m.youtube.com/watch?v=3x1b_S6Qp2Q
[2021-02-13] Lex Friedman Podcast #159 – Richard Craib: WallStreetBets, Numerai, and the Future of Stock Trading - Looks at the recent WalkStreetBets and Gamestop Saga that caused such a stir in the trading markets, then moves on to talk about hedge funds, shorting, new ways crypto and AI could change the world of finance, aswell as more esoteric topics like running a startup and the meaning of life
https://lexfridman.com/richard-craib
I might have a listen to the cringey one you linked to. I wonder if it was just an off day for him or perhaps he didn’t gel well with the guest.
There were a few bits where it got a tad uncomfortable, maybe even cringey (though I think that’s a stretch), sometimes because Lex was being a bit argumentative, or because some of his comments and questions were a bit lame, but there were also some bits where I felt Keller was being a bit prickley.
Overall though I learnt a lot, Keller has a very interesting angle on the tech sector because of where / how he’s been involved.
Even though there were a few places where they clashed a bit, there were also some fun shared humour parts and I felt it ended well. Sounded like they got on quite well despite their differences.
If I have some time (unlikely), I might watch the video version, it would be interesting to see if there’s a big difference in how the interviews come across when there is a video component.
#162 – Jim Keller: The Future of Computing, AI, Life, and Consciousness
https://lexfridman.com/jim-keller-2
I haven’t listened to it yet, the topics covered in the show notes look pretty cool.
Javascript is still the only meaningful "Full Stack" platform and responsible for Billions of dollars in online business and Trillions of dollars in business value.
The Javascript platform is used to capture user behavior in the most efficient manner. This is the fundamental basis for all the online business processes and resulting business value.
Those clicks usually translated to HTTP calls.
What is so fundamental about business processes that they can't do an extra page render and HTTP call? (I imagine there may be some, I just can't see how they are so fundamental and on the order of billions.)
Regardless I don't see how JS is fundamental to any of those services, not even Maps. Instead I suspect HTML would have become more feature rich sooner than HTML5.
I also don't believe JS itself contributes significantly to the trillions of dollars involved in online business.
Anyway, my point is that none of the interactivity as delivered by JS / ActiveX / Flash are themselves reason for all the money in online business. Before graphical browsers there were probably billions in transactions happening on computer networks. And likely would have grown to similar highs without the ability to run untrusted, 3rd party code on demand.
"I do not entertain hypotheticals. The world as it is is vexing enough." - Col. Stonehill, "True Grit"
But the pressure on the Web to be interactive enough vs. fat or old clients, never mind plugins, pushed not only JS but Java into the Netscape browser in 1995. I don't see how that could have gone another way in time to avoid MS doing the same.
The idea of a general Web scripting language was to make a relief valve where developers ran up against the limits of existing markup and less expressive languages the browser processed. Then standards bodies were supposed to study the top JS libs and absorb their Turing-incomplete bits into new HTML and CSS. This happened eventually, but not for so long that we had to invent XUL at Mozilla, Adobe did Flex, etc.
Microsoft - introduced AJAX to support their web enabled servers and online services. Have since morphed into Cloud based company.
Apple - lots of iOS apps are simple webviews that use javascript.
Amazon - online sales and cloud services using javascript.
Google - online ad services using javascript.
Facebook - online user behavior analytics using javascript.
Pretty much the entire online economy has huge dependency on javascript.
One issue I’ve had with Greek philosophy is that although I know there’s interesting content within the stories, it all feels very distant, I can’t pronounce many of the names, a lot of the stories don’t feel super relevant, so there isn’t much that ‘sticks’ in my head.
It’s a bit like reading Shakespear, I rarely get into it, spending all my thought decrypting low level meaning/semantics, and I just never get to the interesting higher level story arcs, or when I do, my flow is too easily broken by some weird old English saying that I am unacquainted with. I guess it’s like any language, lots of practice necessary.
I’ll have to take another look at the Socratic method in more detail. If it’s a technique you are using, based on your interviews that I’ve listened to, then it’s clearly beneficial.
One which I really enjoyed was:
Zen and the Art of Motorcycle Maintenance
https://en.m.wikipedia.org/wiki/Zen_and_the_Art_of_Motorcycl...
I don’t know of many others like it.
I’d really like to ‘get into’ the classics, but it’s like I need to onboard by reading a bunch of contemporary motorcycle style versions first, so I have some familiarity with the concepts.
It got me thinking though, about the value of having some philosophy in comp sci courses.
Back when I was doing my masters (2002-2003) it was a different era for the web, there wasn’t much thought to ethics or philosophy, but I think it could be very relevant these days. It strikes me that it could be a particularly good time to read the ZAMM book.
I don’t know much about comp sci courses these days, maybe they do have such modules included.
Having had a glipse of a CS degree that's part of another school's engineering department, I believe their structure is more rigid, and specifically detailed for 4 years.
https://markjgsmith.substack.com/p/mark-smiths-newsletter-20...
I’d be honoured to have you as a reader, no pressure though, I bet you are super busy, and you probably get a zillion requests like this.
The biggest part of a "good interview" is getting the guest to say "yes" to being interviewed.
The issue that people overlook with "challenging questions" is that the guest must agree to be interviewed in the first place. If the interview is denied, then the "lack of challenging questions" is a moot point because the guest never made himself/herself available to be made uncomfortable.
In other words, let's say Lex starts out of the gate 3 years ago with a hard-hitting style. This instantly increases his rejection rate and decreases his guest count. Then later guests he tries to pitch to also look at those previous challenging interviews which is distasteful to them and they also decline the invitations creating a snowball effect.
You can't replay history with (as of today) 160 podcasts and say they should have been more challenging -- because we wouldn't even have the 160 podcasts to complain about.
Yes, journalists like Kara Swisher like to challenge/ambush/provoke/etc their guests but not everybody wants to be interviewed by her. See the chicken & egg issue?
Lex isn't poor at challenging questions. In reality, he's building a body of work that solidifies his reputation for interviews in a friendly conversational style that doesn't threaten the guest. This reputation attracts many types of guests that a bigger journalist like Kara Swisher can't get.
By building a relationship over time it's more likely that the guest is more comfortable with "challenging questions" (Whatever that truly means).
Charlie Rose, Joe Rogan, Larry King - they didn't really push very hard either.
Sometimes just giving a people a forum to speak works just fine.
Lex has a very odd personality for media - but it works fine for this format, in this industry, and he's smart and curious enough to ask reasonable questions.
He 'talks up' to his interviewees like he's one of their team members of interns and that's fine.
The fact that you would attack me without understanding the thread (and use it as an opportunity to divert people to your Twitter) reflects badly on you.
Noun: a dishonest scheme; a fraud.
"fraud": "wrongful or criminal deception intended to result in financial or personal gain."
Also, as the tweet I linked above said, we didn't profit from either the fast-fixed error of showing unverified creators as being tipable with our token grants to users (we don't make money recirculating our own grants) or the refgate case (we turned off that future-share-of-trading-commission affiliate revenue with Binance).
Hanlon's razor explains these blunders without jumping to scam or fraud, which is a crime we have to fight on the Brave Ads front, so I take it seriously.
Should someone not use Brave on account of these two mistakes? Not for me to say; no browser is perfect. But they were not scams or frauds, and we fixed code and design processes to avoid anything like them in the future.
You went thru various dubious business models before you even got to your current incarnation.
But the most current of your controversies is covered here: https://davidgerard.co.uk/blockchain/2020/06/06/the-brave-we...
And discussed here on HN https://news.ycombinator.com/item?id=23442027
That alone is enough for me to call Brave 'scammy'.
Gerard lies about us "hijacking links", which I'm sure he knows has a well-defined meaning: changing the URL of a link the user clicks on in a page. We've never done any such thing, either.
https://news.ycombinator.com/item?id=25841456
If you want to believe we did these things, you're believing misinformation. If you can't verify these claims are false, I can help; many neutral third parties can help too. But it seems to me you want to believe such misinformation. Why?
When someone creates a browser around token and adtech, and offers to be a gateway between content and ad money, it better has a serious reputation of integrity for me to even start digging into the technical details and see what it's really all about.
Until now, the whole project looked scammy because i only paid a cursory glance at it, because i didn't know who was behind it. Now that i've watched the podcast and understand the whole reasoning behind it actually makes a lot of sense.
Or collecting contributions on behalf of creators without their knowledge.
Why did you write this? Where did you read it? It's false, and you can check this by running Brave, even old versions that are still available. Please check your facts.
On collecting contributions on behalf of creators without their knowledge, the collection was a blunder but the funds were ours, from our token user growth pool. We were not taking funds from others. We fixed this as soon as I caught it, and I'm sorry I didn't catch it sooner.
https://www.computerworld.com/article/3284076/brave-browser-...
Exactly the sort of thing I would expect from a CEO of "Brave".
Not that it's really needed when you know exactly what I"m talking about and are just feigning ignorance. Brave has been called out since their first attempts at finding anything, no matter how scammy, to make money at the expense of the web.
Congrats at finally finding something that sticks - few have done as much damage to the public web as you.
Also, it’s a little ironic that the web sucks mostly due to excessive/ unnecessary JavaScript, so Brendon Eich is responsible for both the disease and the cure. (I say this as someone who actually enjoys writing JavaScript, I should add.)
Brendan Eich is going to be attacked by Google somehow for something, maybe through lobbying in Congress around cryptocurrencies. These ideas are a credible threat to their advertising revenue.
Is their moat as wide as they think it is? They don't exactly have a healthy revenue diversity.
If you like I can say 10 years but that doesn't advance the conversation in any meaningful way.
Is it worth jumping back in for another try?
The MIT AI podcast => AI podcast => Lex Podcast worked out great for him, though...but it feels kind of gross. There are some fascinating guests, but it feels like the content is good in spite of Lex and his “journey”.
None of these podcasts talking critical about cancel culture. There is a far, far better (critical) take on it within "marxist" culture, e.g. from Natalie Wynn. If Lex wants to talk cancel culture, why not invite her for example? She is an academic, smart and wrote and talked extensively about it. But that would be a little too open-minded wouldn't it?
I think you are drawing a false dichotomy though. Somehow, cancel culture is a topic, but they aren’t talking __about__ it. Is there some other preposition that can fill the blank?
Like, how could you sympathize with spaceflight research spending without talking about possible destinations and why you wanna go to space? Spaceflight, like cancel culture, doesn't happen in a vacuum.
Gladly, he isn't still Sam Harris (although the "love" and "beauty" is getting annoying too) or Rogan (lol). There are really interesting episodes, and I overall like the naive questions sometimes.
I think The After-On will restart soon and takes the spot filled by Lex right now.
So now he brings people who study quantum physics, philosophers, thinkers, writers, and he's free to ask more open ended questions.
Might be worth another shot, but maybe just pick the podcasts in the areas you're interested in. I can see how some guests may be a turn off.
As an aside, I wish there was a simple/standard way to link to podcast episodes.
Seems to me the way it should work is that a podcast has its own website with a URL for each episode, but makes the episodes available through many different platforms. The indieweb folks have a term for this sort of thing, POSSE, Publish (on your) Own Site, Syndicate Elsewhere. [0]
Many podcasts upload to YouTube, which naturally supports linking to a specific time in an episode.
Related: it's a pity so many podcasts seem to view it as a their duty to obscure the URLs of the audio files.
- Browser and JS history - Adtech and trying better with Brave - Has SV run it’s course
So we don't have to wonder at all, it would have turned out exactly the way it did and the reason is because things like undefined behavior, zero terminated strings, null pointers, etc have positive utility underappreciated by academics.
If C had been designed as a safer language (with even just small improvements that would not have been hard to implement), it would still be widely used and there would be fewer security problems associated with it. It is the Just World Fallacy [0] to assume that the most widespread language became so on its merits and not based on its historical context.
About Ada, it is actually a decade newer than C. In 1982 I was part of a group that was keeping track of its development and had learned C from old Bell Journals. I was able to a buy a book on Ada and the famous K&R C book and decided I preferred simplicity, so I donated the Ada book and dropped out of the study group. It is ironic that the even more complex (than Ada) C++ was what eventually became popular.
Ada history: https://www.adahome.com/History/
FYI, I worked on a self-hosted Ada compiler as a grad student intern at Rolm (the Mil-Spec computer side, not the digital PBX side) in the summer of 1984.
It also seems to have more powerful ways to set the exact machine representation of types than C.
We booted the ada disk first thing in the morning, changed some writeln() statements, and kicked off the build.
After lunch, the executable was ready (this was not "hello world", but a system with maybe 50K SLOC.
Not practical on that date, with machines that existed. As for writing code that worked, and eliminating defects by language design, ada has no peer. Even in 2021.
Today compile time is not a problem.
That aside, work got done. You spent a little more time checking your code for errors instead of letting the compiler do it.
https://www.goodreads.com/author/quotes/670307.Friedrich_A_H...
Did you listen to Eich’s statements acknowledging the errors and also the hypothesis of what would have been if not JavaScript? Since you have done some accounting, what have been the productivity gains and new frontiers opened by C? What is the amortization of damage over the length of delay?
https://faculty.nps.edu/dedennin/publications/National%20Com...
Redmond is in WA not CA. If your point is about commercial pressure, not just SV bashing, then please listen to my tale of David Hyatt and grad school friend (Hyatt writing draft code and rewriting five times) from the interview with Lex. That was an academic assignment, not commercial work.
The point is not about Silicon Valley, commercialization of the Web (although as Dorothy Denning observed, the pre-Web Academic Internet was a more secure world only because not many attackers yet), or academic pressure. WiB applies to biology. I have friends at a synth-bio startup, one tweeted humorously the other month about how bad a hack evolution is!
He isn’t the best interviewer because he doesn’t have the experience yet but the podcasts are very good because he gives the guests an open stage, he is never adversarial and he never seem to push the interview into a specific direction.
Overall it wouldn’t surprise me if many of the guests would just enjoy to sit down and talk to someone who is without sounding too elitists a smart as they are, being able to reach to a lot of other people is just a bonus.
He's been running the podcast for two and a half years, and this episode makes 160.
How much experience do you reckon it takes to get "good" at interviewing?
Also keep in mind Lex isn’t a professional journalist he isn’t even what one would call a science communicator.
His podcast is still mainly two people having a conversation than a structured interview but he has gotten better at it overall. His early podcasts had quite a few wired moments where you can see that Lex was somewhere else thinking about something these are rarer now.
Your criticism is an example of taking your personal opinion that he has no charisma or interviewing skills and then incorrectly projecting it on to guests that don't share your opinion.
Think about how the guests must voluntarily agree to be interviewed by Lex. Some might go through these steps:
1) get inquiry email from Lex : "Hi, I have a podcast and I'd love to interview you. [... blah blah blah ...] For reference, here are some links to interviews with previous guests [...]"
2) hmmm... who is this Lex guy?!? Let me look at his previous interviews...
3) Ok, he's non-threatening, engaging, and intellectual with his questions. I can sit down and chat with him for few hours.
People oversell the "connections" angle. Yes, he has some connections (Drexel, MIT, ex-contractor at Google, etc) but that just means some warm introductions and a potential guest will politely reply to an email inquiry. Guests can still decline the podcast. Connections don't guarantee the guest will actually agree to talk for 2 hours on camera.
E.g. Chris Lattner has been interviewed by Lex 2 times. Unless he's lying, his comment doesn't seem like he shares your opinion: https://news.ycombinator.com/item?id=24831157
EDIT to reply: "just by cold emailing."
I didn't say he was only sending out cold emails. (I mentioned "warm introductions.")
In any case, the main point of my comment is that you are (inadvertently?) insulting some of his guests. In other words, have you considered the possibility that many of his guests enjoy Lex's interviewing style?
I think the technologist first, interviewer/journalist second is important.
It's crazy to me that people are now claiming that asking "hard questions" is what is ruining journalism. As if we need more of Cuomo interviewing Cuomo on CNN.
As a counterpoint, watch the end of his podcast with Elon Musk, where he made him read the Carl Sagan quote [1]. This was quite brilliant I thought.
Large number of views/listens thanks to Joe Rogan + extremely safe interviewing style to keep the momentum going.
It's really not all that impressive given the absence of alternatives (try searching Wolfram on youtube and see how many other podcasts show up) - it's more surprising how bad the podcast discovery scene is that almost anyone who self promotes through Joe Rogan and puts in the work becomes a top 100 podcast worldwide.
I don't know if you've ever interviewed someone, but that kind of awareness and quick thinking is really really hard.
I wonder how things would’ve turned out differently, had he implemented Scheme instead. Lisp in the browser.
VBScript, the syntax is aesthetically less pleasing, but there is no reason why it couldn't do anything that JavaScript can do, if more verbosely. (Of course, the semantics are inferior as well, but if VBScript had taken off, its semantics would likely have been improved at some point.)
Indeed, if we look at VBScript's cousin, VB.NET, we find both first order functions and async functions, at least in its newer versions.
(Keep in mind that those features weren't present in older versions of JavaScript either. Async functions were only added in ECMAScript 2016. First order functions have been there from the start, but ECMAScript 2015 greatly improved the syntax for anonymous functions with its introduction of arrow syntax.)
I'm genuinely terrified.
In VBScript that runs under Windows Script Host, callbacks are implemented using the WScript.ConnectObject method.
(I'm not sure about VBScript running under IE or classic ASP, if they could use the same method or had to use a different one. But anyway, if VBScript can do it in one context but not another, that would be a deficiency they could have easily fixed if there were a demand.)
Until, US anti-trust investigation and law suit against MS. Which hampered MS enough to open the web business to smaller competitors, like Google, Apple, etc.
The fact that Apple now completely owns iOS and is gatekeeper to mobile app store is ironic. And, Google owns online ad services and is gatekeeper to online ad space is also ironic.
https://thisdayintechhistory.com/05/26/bill-gates-internet-t...
to all the material discovered during US v. Microsoft, we know MS's plan was exactly what you said, after killing Netscape to avoid it becoming the new cross-OS runtime for apps.
Good point about MS leaving the Web fallow, which missed big opportunities Google and Apple (iPhone original app model was Web-only; native came almost a year later for games and for want of WebAssembly :-/).
As I said, in the beginning (Mocha in ten days), == was just as strict if operand types differ: false result. The lesson is to say "no" by default and reflex to RFEs, consider only on repeated request and be willing to find what they truly want that's not a hole in the head ;-).
The other sloppiness in JS is `var obj = {prop:42}; obj.prep` where you get undefined instead of a thrown exception (as you would get in Python). That was due to 10 day rush and little time after for core language work leaving 0 time to add try-catch-finally.
Lloyd suggested `with` after the ten days, also the String.prototype.{anchor,bold,...}() methods, and I added them in the same spirit that I made == sloppy on request from others in the ex-Borland gang of four (Bill Turpin, I think). I should not have agreed to any of these. Lloyd can take credit for them. The String methods are the least offensive but not that useful, now-archaic.
Someone (I think jevering) wrote a dishonest Quora answer claiming Lloyd wrote an engine and I used or benefited from his code. That's false. Mitchell Baker, who was a Netscape lawyer at the time, came on Quora along with others from Netscape management to back up my testimony.
It's funny people want credit for JS when boosting their resumes (Lloyd did great with Looker and as far as I know doesn't need money), while I get blame still. As I should: again, I'm the idiot who agreed to these Borland-flavored requests: sloppy ==, `with`, HTML helper string methods. My blame and credit, mine! But Lloyd is welcome to a share of the blame if he wants. :-D
Thanks, Brendan!