HomeVideos

216 Tasks. 18 Hours. AI Controlled Everything

Now Playing

216 Tasks. 18 Hours. AI Controlled Everything

Transcript

574 segments

0:00

The world's biggest tech billionaires

0:01

complete more in a day than the rest of

0:03

us do in a week. But, is it even

0:05

possible for one of us to be as

0:07

productive as one of them? We decided to

0:09

find out with a totally average person.

0:12

Hey, I

0:15

This is Craig. We set him up with the AI

0:17

workspace GenSpark, which split his day

0:20

into 216

0:22

individual tasks specifically tailored

0:24

to him.

0:25

>> I've got to upload basically my whole

0:27

life. All of my text messages, my

0:29

emails, I I guess also my journal. But,

0:31

will GenSpark be able to help Craig

0:33

complete all 216 tasks? How much more

0:36

time now? The most stressful day of

0:38

Craig's life begins now.

0:44

Hello. This is Elena, your AI assistant.

0:46

You have 216 tasks. I added one more.

0:49

Survive this day.

0:51

For your first tasks, let's start with a

0:53

shower, news and email recap, shaving,

0:55

and vocal warm-ups. Welcome to your

0:58

customized [music]

0:59

daily news briefing.

1:00

>> 847

1:02

unread messages.

1:03

>> Move on to some lip trills or tongue

1:05

trills.

1:05

>> Uh

1:07

I'll state my case, of which I'm 216

1:11

tasks only works if nothing takes more

1:13

than like 5 minutes. So, the more that

1:15

he does concurrently, the more time that

1:17

he banks. And he's going to need that

1:19

extra time as his day gets way, way

1:22

harder. Like to the point where it even

1:24

puts his love life at risk. Don't worry,

1:26

that is going to come back later. By the

1:28

way, I'm Philip, producer from GenSpark

1:30

who's helping oversee everything.

1:33

Hello.

1:34

Oh, hey. Craig started off his day by

1:37

showing me this crazy planner he made.

1:39

>> Uh you see there's literally 216 tasks.

1:43

>> Oh my gosh. So, these are all your

1:45

tasks. That And this is connected to

1:47

GenSpark? [music] Yeah, it is. So, I did

1:49

some research, and it turns out a lot of

1:51

these tech moguls use something called

1:53

5-minute chunking, where they split

1:54

their entire day up into 5-minute

1:56

chunks, completing a different task

1:58

every 5 minutes. If you multiply that

2:00

against the 18 hours a day that they

2:02

work, according to Elon Musk, you get

2:04

216 tasks.

2:07

Okay, so it's looking like the next task

2:09

is uh breakfast. So, could you look in

2:11

the AI drive for the files I uploaded uh

2:14

specifically of what's in my fridge, so

2:15

you can tell me what I can eat? We asked

2:17

Craig to do what most people forget to

2:19

do, which is give the AI tons of

2:21

context. [music] We had him take

2:23

pictures of everything, because in

2:25

practice, the more context the AI gets,

2:27

the more useful it is.

2:28

>> Last night I took pictures of my fridge,

2:31

so hopefully it can come up with

2:32

something.

2:32

>> Checking your AI drive for cabinets,

2:34

refrigerator, and freezer contents. You

2:36

have extremely limited options. However,

2:38

a pickle is a very [music] typical part

2:40

of a Japanese breakfast. So, you are

2:42

having a minimalist Japanese breakfast.

2:44

Enjoy [music] with coffee. Awesome.

2:47

>> [laughter]

2:49

>> Okay, so next task is getting the inbox

2:51

to zero. So, if you could just go ahead

2:53

and look at like all spam and

2:55

promotional material, and just go ahead

2:57

and get rid of those, mark them as read,

2:59

and delete them if you can. I'm now

3:01

working on clearing out the spam and

3:03

promotional emails from your inbox,

3:05

marking them as read. I'm down from 847

3:08

down to nothing, cuz it's all read at

3:09

this point. With emails done, it was

3:12

time for Craig to start work managing

3:13

his dad's pool servicing company.

3:18

Hey, what's up, man? Hey, dude.

3:21

>> [laughter]

3:21

>> Hey, man. I got to show you this like

3:23

you know that welcome packet we've been

3:24

talking about? Uh I do not. It's got

3:28

like

3:29

every definition that a new customer

3:30

might face from us. Like that it

3:32

literally is like a one-stop shop to

3:34

understand everything that we're going

3:34

to be doing. Did you make this? Yeah,

3:36

man. All right, I got to keep pressing

3:39

and pushing on this. I'm doing that task

3:40

thing today, where I've got 5 minutes

3:41

per task, so we're just going to keep

3:43

moving forward. I'm so I'm sorry, Dad.

3:47

Yikes. So, I've been trying to open a

3:49

Shopify store. If you have New York

3:51

Rocks, you paint little faces and like

3:53

cultural icons and stuff like that.

3:55

Craig's next task was to set up his side

3:58

hustle, an online store that sells

4:00

custom-painted rocks. [music] This is

4:02

like actually why I took the challenge,

4:04

is because I was like, oh, like with

4:05

everything going on, like I actually

4:07

will be able to utilize [music]

4:09

AI to get this going. Have you guys ever

4:11

wanted more art in your home? And you

4:12

miss New York City?

4:14

New York Rocks. Definitely not stolen

4:17

from Central Park. $200 to $600 each,

4:19

you know? It's like a collectible thing.

4:20

$200 double? No, it's like a It's a

4:22

collectible thing. It's art. You're

4:23

telling me you don't like the idea.

4:25

Let's get it to build the whole thing.

4:27

Like what the store set up, product

4:30

listing. I also want product

4:31

descriptions and what that would look

4:32

like.

4:33

>> I've set up the agent to generate your

4:34

Shopify store package. This was a huge

4:37

task, so Craig left GenSpark to work on

4:39

it in the background as he moved on to

4:40

something else. And it's found this

4:43

challenge that I gave myself a while

4:44

back to like tweet more. Why can't I

4:47

tweet as much as like these tech

4:48

billionaires, right? I started off by

4:50

tweeting, is soup just a socially

4:52

acceptable drink, or am I missing

4:54

something? It has no retweets or likes,

4:56

so I guess not very good. Want to tweet

4:59

12 times a day. I want to make sure that

5:01

they're funny, so I can go viral.

5:02

>> So, the AI created 12 posts for the

5:05

entire day. Here's the first one.

5:09

After this, Craig worked on a problem

5:11

for his dad's company. We just got a

5:13

full-time tech, like the first one. He

5:15

gets stuck in traffic a lot, so it's

5:17

already identified here that we need to

5:19

optimize his daily route based on

5:20

traffic. So, I think that's what I'll

5:21

focus on. To complete this task, the AI

5:24

found the best route to get to each

5:27

client, laid out the shortest path, and

5:29

then sent it to his dad's employee,

5:31

Peter.

5:32

>> Okay, uh losing time here. Oh my god.

5:34

Okay. Craig spent too much time on this

5:36

task, and he started to fall behind on

5:37

the new ones. Can you go ahead and put

5:39

the real-time traffic data Go ahead and

5:40

generate that financial report with some

5:42

of the top competitors. Can you give me

5:43

the three best workout routines to do if

5:45

I only have 5 minutes?

5:51

All right. While Craig [music] was

5:53

working out, GenSpark kept tweeting for

5:55

him.

5:58

Okay. How much more time now? The task

6:00

was completed.

6:03

At this point, he'd only finished a

6:05

quarter of his tasks and still had 170

6:08

left, which is an impossible number

6:09

unless he picks up the pace.

6:13

Oh, man, I'm getting pretty hungry.

6:15

Oh, uh you you all out of pickles? Yeah,

6:18

I don't know if I could eat another

6:19

pickle.

6:20

>> [laughter]

6:21

>> What if I could see like what should I

6:23

have for lunch based on my last Amazon

6:25

order? Oh, you mean like an Amazon like

6:27

grocery order?

6:27

>> Yeah, like Amazon grocery.

6:29

>> While one GenSpark agent worked on

6:30

getting Craig lunch, another began

6:32

diving into his personal life. It found

6:35

a text thread between my sister and me,

6:37

where she was roasting me about how I'm

6:39

not getting any dates off of my dating

6:41

profile. So, I think I need to switch up

6:43

my picture game.

6:44

>> GenSpark analyzed his profile photo, and

6:46

I helped him take some new ones. I think

6:48

we got some great options.

6:50

But, just when the tasks were finally

6:51

starting to get done, his dad's

6:53

employee, Peter, called with some bad

6:55

news. Hey, Peter, what's up? You're

6:57

kidding.

6:59

The new the traffic map, like the new

7:00

the new address? God, I'm sorry. Okay,

7:02

let me just Just be on standby, and I'll

7:04

I'll figure something out. I'll call you

7:05

right back. Oh, god. Sorry.

7:08

Oh my god, that's so frustrating. Um

7:11

Peter, the tech, like the pool tech,

7:12

he's like at a house with that new

7:16

traffic update and like the report with

7:18

the new customer, and they don't even

7:20

have a pool. I'm here for you. Traffic

7:22

report that we were doing, combined that

7:24

report we made, sent Peter the tech to

7:26

the wrong house. I need like ASAP

7:30

a reconciliation of it.

7:31

>> I'm sending the corrected list of

7:33

results over to Peter now. After this,

7:35

the AI's analysis of Craig's text

7:37

messages called out that he'd been

7:39

losing touch with his sister. So, his

7:42

next task was to improve their

7:43

relationship. I have an idea. My sister

7:46

is Taylor Swift obsessed. Why not create

7:49

a daily routine where you make a phone

7:51

call to my sister, [music]

7:52

and you analyze a YouTube video about

7:56

Taylor Swift lyrics. Just pick one, and

7:58

then read the transcript, and I want you

8:00

to summarize the most fascinating

8:02

details, and then uh call her with like

8:05

uh daily Taylor Swift is a genius

8:06

update.

8:07

>> Taylor Swift lyric update routine is all

8:09

set up. And at the end of the day, Craig

8:11

will get to see his sister's reaction to

8:14

that routine. Craig's next task was to

8:16

draft a reply to a girl he's had on his

8:19

mind. How long have you been talking to

8:22

Ali?

8:22

>> Ali and I have probably been talking for

8:24

2 weeks. Can you just

8:27

make the plans with Ali as you see fit?

8:31

Saying it can draft a message, and I

8:32

just need to approve it. I can [music]

8:33

send it over. Hey, exclamation point.

8:36

I would love to hang out. No punctuation

8:39

after the second one. We have like no

8:40

time left. I'm I'm just I'm just going

8:42

to send it.

8:42

>> It didn't tell me who I'm going to be.

8:44

Like it didn't command that from me.

8:46

Instead, it like took pieces of me to

8:48

build up what I actually want to be, and

8:50

then allow me to be the own architect of

8:52

that. She just hearted my message and

8:54

said tonight. Tonight?

8:56

No, I I can't I've literally been trying

8:58

to go on a date with this girl for 2

9:00

weeks. The AI closed this in two

9:02

messages.

9:04

Are you saying it has like better game

9:05

than you? Unfortunately, it has better

9:07

game than me.

9:08

>> [laughter]

9:09

>> She just said we can do something

9:11

casual, like dinner and a movie or

9:13

something. And I said, yeah, sure, why

9:14

not casual, for sure.

9:15

>> Create a menu plan with recipes for

9:17

homemade pizza. One of the options is it

9:19

can call local chefs a private

9:22

Wait. Uh start searching for private

9:23

chefs in the area with availability for

9:25

tonight, and seek [music] pricing for

9:27

approval from me. Understood. We're

9:29

pushing forward right now to get the

9:31

private chef to your location as quickly

9:33

as possible. That is crazy.

9:35

>> That's wild. What other options do I

9:36

have? I need to get all these tasks.

9:38

Like I need to get this going now. While

9:40

GenSpark worked on more than 20 tasks in

9:42

the background, Craig decided to finish

9:45

a few more on his way to the coffee

9:46

shop. Hey, do you remember that novel

9:48

that we were writing, how we can

9:50

increase that tension to keep the

9:52

readers' attention,

9:53

so we can actually get them into chapter

9:55

four? At this point, GenSpark was

9:57

working on so many things that Craig's

9:59

main task became just making sure no

10:01

mistakes were being made. Hey, so I got

10:03

this 2012 Nissan Sentra and my

10:05

registration has been out and it's due

10:07

like yesterday's. Could you find a

10:09

mechanic in my area of Brooklyn, call

10:11

around to get the best quote just so I

10:13

know I'm, you know, not getting screwed

10:15

over? Near car registration? Hi, I'm

10:17

Alina, an AI assistant calling for

10:19

Craig. GenSpark wasn't just

10:20

multitasking, it was multiplying and

10:23

Craig was impressed. He was on three

10:25

separate phone calls at one time. Please

10:27

leave a voicemail with your name, phone

10:29

number, and a brief summary of what we

10:31

can help you with. Hi, this is Alina

10:33

calling on behalf of Craig. He needs a

10:34

quote for replacing the catalytic

10:36

converter on a 2012 Nissan Sentra. Even

10:38

if I could do that, I've only got one

10:40

phone. Do you know what I mean? So, that

10:42

was like incredible. Perfect. So, that's

10:44

a success.

10:50

There's a bag on the ground. Oh, no.

10:52

Look at literally.

10:55

Chicken noodle? Some Red Bulls? You want

10:57

one? I guess this is lunch.

10:59

Yeah. That's lunch. All right, what's

11:01

next? Go ahead and schedule a feedback

11:03

call with my dad concerning the traffic

11:06

routes.

11:07

I mean, I'm just looking at the math

11:09

here. It doesn't seem like you're very

11:10

on track. Is someone knocking?

11:14

Craig?

11:16

Hey, but what's your name? Saptar, nice

11:19

to meet you.

11:19

>> Saptar, nice to meet you, man. Yeah, are

11:21

you cook you dinner. So, this AI agent

11:23

called me. I was set up set up and told

11:25

to come cook dinner for you guys and

11:27

here I am. Looks like you got a date. I

11:29

got to get like going on this stuff, but

11:31

um While the chef started cooking, Craig

11:33

continued to knock out [music] tasks.

11:34

Let's get an update. Your Shopify store

11:37

is already ready and your product is

11:39

ready to be sold. Did you guys hear

11:41

that? New York Rocks? Up and running?

11:43

That is like a huge goal win for me, I

11:45

think.

11:46

Nice. Is this your first Shopify store?

11:48

I had the idea for years and here it is.

11:51

It's all like figured out. You have like

11:52

sales tax figured out and all that?

11:55

Oh, god. I've set up a business. Can you

11:58

help me out and figure out how to get

11:59

this thing registered legally cuz I

12:01

guess I don't need to pay

12:03

uh tax. Task to adjust the business

12:05

registration to New York Rocks LLC. Each

12:08

task started taking Craig longer and

12:09

longer [music] and everything slowly

12:11

started to turn into chaos. It does feel

12:13

like you're spending a lot longer than

12:15

five minutes on this. [music] Yeah,

12:17

you're right. You're right. You're

12:18

right. Uh the the competitor analysis in

12:20

the art space. Can we look back at that?

12:23

But I mean I mean for sorry. So, I mean

12:25

filing status

12:27

No, I

12:29

No, no, no.

12:33

The toy industry competitor SWOT

12:35

analysis includes a thorough breakdown

12:37

of strengths,

12:39

The chef needed help rearranging things,

12:41

distracting Craig even more.

12:44

Like go long long list The agent is

12:46

currently working on So, is there is

12:50

Oh, you got the paper towels yet?

12:52

I'm gathering options for purchasing

12:54

more paper towels, including prices,

12:56

>> [music]

12:56

>> quantities, and delivery details.

13:05

This is my business and it keeps my

13:07

personal money and my business money

13:09

separate.

13:17

>> [music]

13:20

>> What does everyone think? Looking good.

13:22

>> Yeah, you think? I don't have a strong

13:24

opinion. Oh, okay. [laughter] Just got

13:26

an entire order of paper towels here

13:28

from earlier. Dude, this is insane, man.

13:30

Like seriously. Oh, thank you so much.

13:32

I'll put it on the table. Thank you.

13:34

Thank you. Yeah. It was almost time for

13:35

the date, but the chef wasn't done yet.

13:38

>> Oh, perfect. Yeah.

13:39

>> [laughter]

13:39

>> Yeah, man. Look at that. Oh my gosh. So,

13:41

Craig did what he had to do. Let's go

13:43

here. Yeah, absolutely. And it's

13:45

actually going to be a lot faster if you

13:46

just go right. If you just go right out

13:48

just right out this way then

13:51

Yeah, perfect. He's parked out there.

13:53

>> You literally sent him the opposite

13:55

direction. But that's the direction that

13:56

she's coming from. If so, they would

13:58

cross paths.

14:01

>> [laughter]

14:03

>> While Craig finished setting the table,

14:05

I went out to go meet his date.

14:07

>> Hi. Do you know who Craig is? We've been

14:10

[music] talking a bit, just kind of see

14:11

if we had like a connection in person.

14:13

What do you know about him? Um

14:16

I'm going ahead and AI being

14:19

retrospective on everything that we've

14:20

been going over. Even right now, I've

14:22

got an additional 36 tasks running

14:25

throughout this dinner.

14:26

>> We weren't like steady communicating all

14:28

day every day.

14:29

Um and then just kind of out of nowhere,

14:31

he

14:32

he kind of took that initiative. Hello.

14:35

Hi Craig, I'm Ali.

14:39

>> [music]

14:39

>> Dinner was ready, so I left them alone

14:42

on their date while the AI actually

14:44

continued working on everything in the

14:46

background. In retrospect, it basically

14:49

sent like three people to my place. That

14:53

is a total game changer. Cheers.

14:56

Cheers. It was offering [clears throat]

14:58

to like analyze her Instagram to give me

15:00

questions to ask like that I was like

15:02

that nervous. But I I don't know. I

15:03

decided to like pump the brakes on that

15:05

one and just like, you know, I I'm

15:07

appreciative of the fact that it helped

15:08

me schedule that and find the time, but

15:10

you know, I didn't necessarily want to

15:12

go all in on that.

15:14

I'm sure you're probably exhausted.

15:16

You've been probably going for a long

15:17

time. Uh well, you know. It was really

15:19

fun. We had a really good conversation.

15:21

Food was very good and the wine flowed.

15:23

[music] We killed the bottle. He's like

15:25

kind of like cute little proud to be

15:27

nerdy [music] type. I was honestly

15:29

really impressed. This was like honestly

15:31

really fun. Like I'm happy that I did

15:33

it. Okay. No, I really liked meeting

15:34

him. It was it was Cool. very good time.

15:37

Sounds good. Thank you.

15:39

Dude,

15:40

I Let's go. I think that was great.

15:42

Honestly, I think that was like so good.

15:46

Craig still had 40 tasks [music] left

15:48

and needed to hurry to finish everything

15:50

in time. Moment of truth. I don't know

15:53

that I think Elon Musk does 216 tasks in

15:57

one day. Well, I mean like the thing is

15:58

like Elon Musk, right? He's like an

16:00

executive of people. He has an entire

16:02

[music] team of people that he's

16:03

directing what to do. I don't have that

16:05

in my life. This actually unlocked a

16:07

level of that where I was able to

16:09

basically direct the agents. So, I ended

16:11

up not having to do the work myself.

16:12

>> I I I don't know, man. It's it's so

16:14

messy after the date. I

16:16

This clean up, it can be one of the 216,

16:18

right?

16:20

I mean,

16:22

Did you end up saying that you made all

16:24

of this? I doubled down and I said that

16:26

um the pasta's handmade. It's very good.

16:28

I made this. She goes, "Oh, right. Yeah,

16:30

you made this."

16:31

Yeah, it would have been so good. Oh,

16:33

really? [laughter]

16:34

How do you live alone?

16:37

No, no, no. Never mind.

16:39

I think it's obvious why you live alone.

16:41

>> [laughter]

16:43

>> My my mom just texted me. She said,

16:45

"Thanks for the kind message." No

16:46

[laughter] way. I had all those emails

16:48

that I was trying to approve.

16:49

>> Hey, go ahead and find my highest

16:52

priority response emails. Anything that

16:53

looks pressing and then if you could

16:55

also construct an a response template

16:57

for me, I would like to use that to

16:59

approve that send to people that haven't

17:00

responded to in a while.

17:01

>> And I guess like my mom weeks ago sent

17:03

me "I love and miss you." As I was going

17:05

down approve approve approve approve, I

17:07

approved a response today that said "I

17:09

love you, too." So, like literally

17:10

that's what she was calling about. That

17:12

is so crazy. [laughter]

17:13

There's some good and some bad, right? I

17:16

think like immediately realizing I'm

17:17

neglecting my own mother is not very

17:19

good, but like the fact that

17:22

I don't know. It took the initiative and

17:24

made her feel great and like hearing her

17:27

voicemail and she was happy about that.

17:28

So, like from one instance, that's

17:30

great, you know, but more of a life

17:32

lesson that maybe I need to be spending

17:33

more time with my own mother.

17:36

Meanwhile, the AI kept posting on X for

17:39

Craig. It's posted eight [music] since

17:41

we started. It's about to post the

17:43

ninth.

17:43

>> I gained two followers.

17:45

Isn't that crazy? And as things wound

17:47

down, GenSpark sent Craig a recording of

17:50

the Taylor Swift call with his sister.

17:52

My [laughter] sister just texted me that

17:54

she said that that was wild with the

17:56

crying laughing emoji. Hold on, there's

17:58

a breakdown. I can I can actually listen

17:59

to it. Today's insight, a

18:07

and

18:08

connecting [music] modern pop with

18:09

timeless poetry. Cool, right?

18:12

That's pretty cool.

18:13

>> [laughter]

18:14

>> Thanks for your time, Kristen. Take

18:15

care.

18:15

>> Another approval for an additional 12

18:17

tweets

18:19

uh scheduled. Amazon item search uh for

18:22

>> It's 216 tasks, which is insane. It

18:25

sounds insane.

18:26

>> Those tasks aren't the defining thing of

18:29

what [music] Craig is, the concept of

18:31

Craig. Since more than half of Craig's

18:33

tasks were completed by GenSpark, it

18:35

left us with a big question. If the AI

18:37

did all the tasks that you could

18:39

possibly do, is the AI actually Craig?

18:41

Or or

18:42

>> [music]

18:42

>> what's or are you Craig? Who's Who's the

18:45

real Craig? I think I'm the real Craig.

18:46

It was doing things that I was directing

18:48

it to do and things that I want to do

18:50

accomplish. But at the end of the day,

18:51

like I was architecting the entire

18:53

thing. It was coming from, you know, my

18:54

mind and what I want to do accomplish

18:56

and like [music] allowing me to better

18:58

live and service my life. I'm the one

19:00

writing my book. I'm the one that's

19:02

interested in singing. I'm the one that

19:04

wanted to go on a date and, you know,

19:05

reconnect [music] with my sister. And

19:07

this was just getting rid of those tasks

19:09

that didn't allow me to better service

19:11

those aspects of my life. Five-minute

19:13

chunking is stupid. Don't do it. That is

19:15

not the good methodology. Don't take

19:17

that from this.

Interactive Summary

The video follows an experiment where a person named Craig attempts to complete 216 tasks in a single day, inspired by the '5-minute chunking' method allegedly used by tech billionaires. Assisted by the AI workspace GenSpark, Craig manages his side hustle, personal relationships, and professional obligations. While the AI successfully automates many tasks—ranging from scheduling a date and finding a private chef to sending emails—it also highlights the potential pitfalls of over-automation and the importance of human intent. Ultimately, Craig concludes that while AI can efficiently handle administrative burdens to free up time, the '5-minute chunking' method is impractical and that he remains the essential architect of his own life.

Suggested questions

4 ready-made prompts