How Silicon Valley Broke Productivity

00:00:00.000 | All right, let's do some questions.

00:00:01.560 | What do we got?

00:00:02.640 | - All right, first question from Sam.

00:00:05.400 | There's a time management popular app

00:00:07.680 | that insists that people should automate their time,

00:00:10.940 | time blocking using AI driven apps.

00:00:13.560 | They think that it's too hard for people to manually replan

00:00:16.280 | and prioritize work.

00:00:17.860 | What's your response?

00:00:19.400 | - Negative.

00:00:20.240 | I do not believe we need AI driven time blocking

00:00:24.720 | because we have to shave off the moments required

00:00:28.440 | to figure out what should be on our plan

00:00:31.720 | or to fix our plan when it actually drifts away.

00:00:36.720 | Now, Silicon Valley is desperate

00:00:40.600 | for knowledge worker productivity,

00:00:42.360 | which is this sort of trillion dollar a year

00:00:44.360 | business opportunity to be something that can only be fixed

00:00:46.840 | by proprietary high-tech tools

00:00:48.280 | that only they know how to make.

00:00:49.360 | There's huge money on the table there.

00:00:51.840 | The unfortunate reality for Silicon Valley, however,

00:00:55.380 | is that this is not an issue that needs complicated,

00:00:59.460 | high technology to solve.

00:01:01.260 | Now, there's a backstory here.

00:01:03.680 | This is a theory I have.

00:01:05.660 | I'll share it with you.

00:01:06.480 | It's my theory about how Silicon Valley

00:01:08.960 | accidentally polluted our understanding of productivity

00:01:13.180 | in a way that I think was very detrimental

00:01:15.580 | that we're only sort of pulling our way out of now

00:01:17.500 | in the last few years.

00:01:18.400 | Here's what I think happened.

00:01:20.180 | In the early 90s, so the 1990s, let's say.

00:01:23.260 | In the 1990s, that whole decade,

00:01:24.980 | when Silicon Valley was exploding into prominence

00:01:27.520 | and into its sort of economic relevance,

00:01:30.980 | this was the same time

00:01:32.580 | when the computer processor wars were unfolding.

00:01:36.060 | So you remember this, this March from the 286

00:01:38.540 | to the 386 to the 486 to the Pentium,

00:01:41.220 | you actually knew how many megahertz your processor was,

00:01:45.580 | and this processor has more megahertz than that processor.

00:01:48.800 | So there's this age of the processor war.

00:01:51.220 | During that age, as Silicon Valley was coming

00:01:53.500 | to cultural relevance and economic relevance,

00:01:57.140 | I believe they adopted implicitly

00:02:00.580 | this computer processor metaphor

00:02:02.260 | for understanding human productivity.

00:02:04.780 | So when you're trying to make

00:02:06.720 | your computer processor better,

00:02:09.820 | what does it mean for a computer processor

00:02:11.180 | to be more productive?

00:02:12.700 | It goes through instructions faster, as fast as possible.

00:02:17.100 | Let's reduce the time and friction

00:02:18.720 | in between each next instruction that we execute.

00:02:22.840 | The other key component

00:02:24.260 | to making a computer processor very productive

00:02:26.160 | was to make sure that it always has something to do.

00:02:29.540 | And so there's a technology called predictive pipelining,

00:02:32.620 | where essentially what would happen

00:02:33.820 | is you're looking ahead to try to queue up instructions

00:02:37.980 | you think are gonna come next,

00:02:39.660 | because you don't want there to be too much downtime

00:02:42.140 | after the processor completes an instruction

00:02:44.260 | to go figure out what it should do next.

00:02:47.060 | That's all cycles

00:02:47.900 | that could have been doing something productive.

00:02:49.640 | So there's this mindset of you want a queue

00:02:51.140 | always full of things to do,

00:02:52.300 | so it always has something to pull in,

00:02:53.940 | so that it's always executing something.

00:02:55.820 | And every single cycle is executing something,

00:02:57.820 | and we want the speed between those cycles

00:02:59.220 | to be as fast as possible.

00:03:01.040 | That's productivity for a computer processor.

00:03:03.180 | Silicon Valley adopted a similar model

00:03:05.620 | for their human employees.

00:03:07.140 | A notion of human productivity that was built around

00:03:11.100 | how do we reduce the friction and time

00:03:13.120 | in between actual tasks or things being completed.

00:03:17.900 | So the focus went into how do we make networks faster?

00:03:20.980 | How do we make email more seamless?

00:03:22.940 | So you can get to the thing quicker,

00:03:24.500 | fewer keystrokes to send that email

00:03:26.100 | and have another one come in.

00:03:27.600 | How can we build information management systems

00:03:29.820 | to make sure that every bit of information you need

00:03:31.620 | is right there at your fingertips?

00:03:33.620 | The notion of computer inboxes,

00:03:36.020 | email inboxes overflowing was not an issue

00:03:38.220 | because like a predictive pipeline,

00:03:40.280 | you would want more than enough stuff always ready to go

00:03:43.500 | so that that human always has something they can do,

00:03:45.380 | an email they can respond to,

00:03:46.540 | or something they can attach,

00:03:48.040 | something they can drag over into this program

00:03:49.620 | and then send that through email to that program

00:03:51.380 | that gets loaded here and it gets put on the screen

00:03:53.260 | and people can see it.

00:03:54.180 | It's all about reducing friction,

00:03:55.380 | increasing the velocity of information,

00:03:57.220 | increasing the velocity of task execution.

00:03:59.820 | It's a very computer processor type metaphor.

00:04:02.460 | And because Silicon Valley became so powerful

00:04:04.700 | and economically relevant in the 1990s,

00:04:06.780 | what they were doing spread.

00:04:08.280 | And we know that happens.

00:04:10.900 | I think that the clearest example of Silicon Valley nonsense

00:04:15.900 | spreading nationwide,

00:04:17.460 | the one that we all know is open offices.

00:04:19.500 | Silicon Valley started doing these open office plans

00:04:22.560 | because in their highly rarefied world,

00:04:26.100 | it really mattered to them that they could signal

00:04:28.200 | the potential employees and potential investors

00:04:30.540 | that they were disruptive

00:04:31.500 | and they were doing business in a new way.

00:04:33.340 | It didn't really matter how they signaled this.

00:04:35.220 | They just had the signal that they were disruptive

00:04:36.740 | because they would get better talent

00:04:37.900 | and they would get more investment.

00:04:38.800 | And that was critical to their survival.

00:04:40.420 | I mean, they could have done almost anything here.

00:04:43.720 | They could have all worn weird, silly hats, whatever,

00:04:46.100 | but they just had the signal,

00:04:47.260 | we're being disruptive.

00:04:48.260 | And then you fast forward 10 years later,

00:04:50.140 | and I gave a talk at a major drug manufacturing

00:04:53.820 | a few years ago, and they were all shaking their heads

00:04:56.100 | about their giant open office.

00:04:57.500 | It made no sense why they had open office, right?

00:04:59.260 | So stuff comes out of Silicon Valley.

00:05:00.820 | So I think this notion of productivity

00:05:02.440 | as computer processor style, picking up the speed

00:05:06.140 | and reducing the friction required to execute small things,

00:05:08.420 | that just spread.

00:05:09.380 | And work and productivity in the knowledge sector became,

00:05:14.080 | are we on it?

00:05:14.920 | Are we quick?

00:05:15.740 | Are you here?

00:05:16.580 | Email's not fast enough, let's do Slack.

00:05:19.080 | Let's do meetings and video

00:05:21.780 | because we can get onto those faster.

00:05:23.780 | How about you just directly have access to my calendar

00:05:26.100 | and can just start throwing things on there.

00:05:27.340 | So it's a speed.

00:05:28.500 | Now, of course, this didn't work at all

00:05:30.980 | because human beings are not computer processors.

00:05:34.980 | We can only focus on one thing at a time.

00:05:36.920 | It takes us a while to actually get going on something.

00:05:39.100 | And once we're done with something,

00:05:40.020 | we need time to wind that back down,

00:05:41.960 | to rest and recharge, and then move our mind

00:05:44.400 | into a new context to work on something else.

00:05:46.320 | I think there's probably like four different things

00:05:48.100 | we could productively give time to

00:05:49.420 | in a typical eight hour day with sufficient rest.

00:05:52.220 | Our brain can't jump back and forth

00:05:53.980 | like a computer processor.

00:05:55.160 | It's not agnostic to op codes.

00:05:57.420 | The thing we just operated makes a big difference

00:05:59.220 | on the thing that comes next.

00:06:00.460 | We're not just circuits being driven

00:06:03.420 | by a crystal oscillator at a constant speed.

00:06:06.300 | So this computer processor notion of productivity,

00:06:08.520 | I think was devastating.

00:06:10.220 | It's a lot of the exhaustion that people,

00:06:12.360 | like we talked about earlier in the show,

00:06:14.140 | Ginny O'Dell, Berkman, McEwen, me,

00:06:18.880 | we're picking up on the exhaustion of this overload.

00:06:21.240 | This overload is in part a direct effect

00:06:23.120 | of this broken model of productivity.

00:06:26.200 | But again, it's not mustache twirling exploitation.

00:06:30.840 | It's, hey, the cool kids are doing this.

00:06:32.960 | Jim Clark just built this giant Hyperion yacht.

00:06:39.360 | If you don't know what I'm talking about,

00:06:40.440 | read Michael Lewis's book,

00:06:42.240 | The New New Thing about the excesses of Silicon Valley

00:06:45.100 | in the 1990s.

00:06:46.940 | They say, so whatever they're doing must make sense.

00:06:48.660 | Let's be more like them.

00:06:49.720 | And they built the tools that we used.

00:06:51.680 | We used their tools.

00:06:53.060 | We tried to emulate how they worked.

00:06:54.460 | That's what I think is broken.

00:06:56.740 | And so no, we're not gonna fix our way out of this

00:06:58.300 | by making those tools faster.

00:06:59.740 | Using AI to manage our time block schedule

00:07:04.800 | is perpetuating the computer processor metaphor

00:07:08.240 | of increasing speed and reducing friction

00:07:10.160 | of task executions to key to productivity.

00:07:12.060 | That's an entirely broken metaphor.

00:07:13.820 | Our issue is not that it takes us too much time

00:07:16.060 | to build our plan,

00:07:18.040 | or that it takes us too much time to change our plan.

00:07:20.820 | The issue is that we have 5X too many things in that plan.

00:07:23.720 | It takes me five minutes to really think through

00:07:26.740 | how to build my plan.

00:07:27.660 | It takes me three minutes to fix it.

00:07:29.060 | That's not the problem.

00:07:30.580 | The problem is checking the inbox once every one minute.

00:07:33.100 | The problem is having seven meetings per day

00:07:35.480 | where you're trying to scramble in between these meetings

00:07:37.140 | to try to answer Slack messages.

00:07:39.040 | That's where the real problem is.

00:07:40.220 | And the problem for Silicon Valley

00:07:42.100 | is that the solutions to that problem

00:07:43.540 | have more to do with getting away from their ideas

00:07:46.580 | and getting away from their tools

00:07:47.820 | than they do about embracing even more.

00:07:50.120 | So, nope, I'm not a big believer

00:07:52.820 | in an AI-driven time-blocking app.

00:07:55.580 | I think my paper planner probably works just fine.

00:07:58.080 | Not that I have a rant or anything, Jesse.

00:08:01.740 | - That was a good rant.

00:08:02.580 | - Not that I've thought about that.

00:08:05.180 | (upbeat music)

00:08:07.760 | (upbeat music)

How Silicon Valley Broke Productivity

Chapters