This Week in Startups: AI Demos! An AI-driven ownership platform for founders, AI interviewer, & more! | E1837

They do have a video ideas.

And so I just dropped in is like, you know, one of the features

that were kind of deeper in the product.

So you give it your URL.

OK. And it gives you a video idea.

And I guess decoding Tesla's energy plant.

Wow. With Elon Musk.

Oh, OK. So just.

Oh, what a great idea.

Get Elon Musk.

Yeah, I was going to say, that's a little obvious, right?

OK. Who's next?

Yeah. Wow.

Well done.

Brilliant. Great.

Wow. Hey, Nick, let's do that.

Yeah. Hey, Nick, we might get we might get some views.

Yeah. I don't know page and Sergey next.

This week in startups is brought to you by a rising

ventures is a holding company that acquires tech startups

facing setbacks.

A rising ventures knows what founders care about

because they aren't bankers.

They are tech founders themselves.

Go to a rising ventures dot com slash twist today

to learn more and connect with the team.

Nuts dot com is your one stop shop for the highest

quality foods for business.

They offer delicious office snacks,

corporate gifts and wholesale ingredients.

Nuts dot com is offering new business customers

a free gift with purchase and free shipping

on orders of one hundred and twenty five dollars or more

at nuts dot com slash twist.

And brave is an internet privacy company on a mission

to protect your personal info online.

Download brave today at brave dot com slash twist

to browse faster, search privately and so much more

all in a single click.

All right, everybody, welcome back to this week in startups.

It's Madra Monday.

My friend, Sonny Madra is here from definitive intelligence

and we do our demo or die every Monday here on this week

in startups, except when I was in the region in the as we say

I came up with a new slang term in the because everybody

well, everybody says, oh, are you in the region?

And I say, yes, I'm in the region.

And then I was like, you know what, just say in the I NDA

and it will be implied that it's the region.

And so you tweeted that and I didn't get.

Yeah. And I'm trying to make it slide now.

So if like, I feel like you were in India when you tweeted that, Jason.

Yeah, it's much more subtle, much more subtle.

But I mean, what a trip.

The stuff going on over there, the investment in AI, speaking of AI

and the investment in startups, whoo, they are.

I mean, those societies, you know, we have all kinds of perceptions about them.

And they're changing month by month.

Now, you know, people are telling me like, if you were here six months ago,

this would have been different.

If you were here 12 months ago, this would have been different.

It's pretty, I have to say, I would dare.

I use the word inspiring to see this rapid change.

Movie theaters, women in all the meetings and music playing.

I was in one place.

There were cocktails.

You know, just things are moving at a very rapid rate.

And the commitment to startups is like nothing I've ever seen.

Like, OK, they have the startup bug.

Like, I don't even know how to explain it.

Like they're into it like people in the U.S.

are into like fantasy football or, you know, DraftKings or something.

Like it is a, it's like a cultural phenomenon.

It's pervasive.

Everybody wants to, you know, invest in startups, start companies.

And it's very interesting.

They don't say investment, they say ticket.

So what is your ticket size?

Oh, and I was like, I'm sorry, cricket, ticket, what?

And it's a ticket size.

So if you were to be in a launch fund for or you were investing in a company

that you would they would just say, oh, well, what ticket size would you like?

It's like a very nice gracious way of saying, well, I want to put $25,000

into that company. Very interesting.

OK. And J. Cal, just for my education, top three companies in the space,

like established ones so that, you know, we can get people like kind of our,

you know, the entrepreneurs look up towards, right?

Like, you know, we do.

Yeah. So the one you'll know is Kareem, which was bought by Uber.

And, you know, I met the founder, C-A-R-E-E-M.

It's like a super app.

And so it was bought by Uber and then the Uber, I think, sold back

like the non Uber parts of it in the name.

And then there's like a buy now pay later company that's doing phenomenally well.

That hit a billion dollar market cap.

So there's a lot of companies, I wouldn't say exactly copycats,

but, you know, regionalized startups that just haven't gotten there.

And it turns out like buy now pay later in the Kingdom of Saudi Arabia or UAE.

Like pretty great. These are incredible consumers.

The arpu of these consumers is bonkers, bonkers.

And like, I'm like, well, if they're rich, like, do they need to do four easy installments?

Like, oh, well, you get a whatever you're going to get.

You can, you know, stretch a little bit and get something else slightly better.

So just like you might in America, you know.

I have a great story with the Kareem founder.

I was at a conference once and we got put in a group together on this hike.

That was like really insane in like the south of Germany.

And, you know, basically it was him and I for like three hours together.

So we really had a good time.

Was it Magnus?

No, no, the Kareem founder, I believe Moodes here.

There's Moodes here.

And then there's Magnus.

I know Magnus. Yeah. Yeah.

Oh, you want to hike with me? That's great.

Yeah, it was awesome.

They have like some things that they've done that, you know,

you don't have an Uber and Lyft here.

So when I opened the Kareem app when I was in UAE in Dubai,

it had Clean My House.

So you can press a button and have a cleaning person come to your house.

So I was like, wow, that's kind of interesting.

And supposedly, yeah.

Supposedly Uber is going to have this isn't any inside information, obviously.

There was a story I saw in Apple News just that they're testing

some sort of task gravity kind of service.

I don't have any confirmation of that.

I just saw the news headline go by.

But let's get into AI.

I have to tell you off the bat, I have started to get great utility

out of ChatGP T4's app because of multimodal mode.

I will share on the screen right now.

I've been getting back on my Peloton.

You guys know I'm on a health kick.

And so I'm trying to do like a zone to walking every day.

So I just took a picture of the Peloton screen

because I've been trying to get my information out of the Peloton.

It's kind of like a locked up system.

It'll work with Apple, whatever.

But I like to just put it in a Google Sheet, right?

I like to track my stuff in a Google Sheet.

Are you doing that like 30 minutes, 10 incline, three speed?

I heard about this 30, 10, 5 or something.

I don't know what it is, but I think it's like 30, 10, 3 or something like that.

Right? Yeah.

Three miles per hour for 30 minutes at 10 incline or something.

Is that what you're doing?

So I heard that this is what all the women in Hollywood are doing.

This is some sort of way to get then.

And I did literally somebody explained it to me as such.

Is that what you're talking about?

This Hollywood lose weight thing?

I don't know. It's Hollywood.

I think like even I see like these, you know, fitness experts on on Instagram

doing it, like they're all over the place.

Yeah. So Peter Tilla, who I read his book that

or I got them like two, there's a way through it.

He just says you should be in zone two.

But I also like interval training because it kind of like sprinting.

So it's kind of hard for me to get on the treadmill.

And I love the Peloton tread.

But, you know, I just getting the information out of there, like it's kind of a

lock box. They don't have like really good interoperability.

The software doesn't get updated.

So I just took a picture of the screen here.

Yeah.

And I'm just doing this very gingerly walk in the mornings.

Is here at 7 a.m. today.

And I did 48 minutes and the workout was running on the tread is really walking.

But if you scroll down, it just takes everything off the screen.

Just I literally took a picture of the screen and I just asked it to

summarize my workout and it gave me my zones and how long I was in it.

Well, then I was like, oh, wait a second.

I had pictures of the previous two days.

So I uploaded those.

And then I put all of them in a table and then I made a second table.

I said, and then average it.

You love these tables.

You're like, I'm a table guy.

I'm a table guy.

Always have been.

Always will be.

Um, I like a little structure in my life.

Sorry.

I like the grid system.

I lived in Manhattan.

You know, listen, I like the grid system.

What can I say?

You know, it's probably just numbers.

You can see the differences.

Wow.

This is.

Yeah.

And I said just to average it or whatever.

And you know, what I'm trying to do is get that zone two.

So you look at zone two, 61% of my time in the treadmill zone two.

26 is zone one.

Yeah.

Um, where like some people say you should be in zone one because

it doesn't spike your cortisol, but zone two does do a little cortisol,

but also, you know, gives you a decent heart rate.

You get to like 65%, 61% of your heart rate or whatever it is.

And so that's where I want to live.

Yeah.

And then I got a little bit in zone three.

Cause it's very hard to stay in one zone.

You know, you might tip up or below.

So anyway, I just thought this was like really indicative of it.

And then I had my colonoscopy.

Everything turned out great.

I had some like one or two little polyps.

They took off not to be like too detailed here, but they sent me a letter

like telling me I'm a okay.

I know you've been asking, uh, I know Uranus is a big joke that we don't

tell on all in anymore.

I heard some people's feelings, but, uh, I just took a picture of the letter.

Okay.

And I said transcribe this for me.

Yeah.

And then I realized, wait a second.

I'm using chat GPT for like I use notion or like I use, you know, the notes app

or I used to use ever know just a way to kind of, you know, whatever.

And I know there's OCR apps there, but I was shocked at how accurate

this thing was.

And then I said, and summarize this for me.

Okay.

And summarize it perfectly.

So I was like, okay, wait a second.

This is going to be my new thing.

Just taking out the app, taking pictures of stuff transcribed and then shred

the document, right?

Get rid of it.

Um, and that is just such a game changer.

And then I get to my desktop.

It's there.

And then it is going to be in my brain, my AI brain for all time, which is pretty

great.

You know, now it knows my workout history.

Right.

And it knows my colonoscopy.

And, you know, maybe I'll do an interview or a zoom with somebody and take

the transcript and put that into chat GPT for her.

So I'm starting to see, okay, it's structuring the data.

It's got it there.

Now I can ask it something.

And then I took a picture of my breakfast today.

Okay.

And give me the calorie count.

Wow.

And it almost nailed what I had in the plate.

It was like half an avocado, some cherry tomatoes, and it was like a two egg

omelet.

I was like, it's a four egg omelet.

It was like, I'm so sorry, sir.

Please don't turn me off.

Yeah.

Um, but it, it's kind of interesting, like I've seen a lot of like single

purpose apps doing these things, but now I'm starting to see the grand vision

here of, Hey, if you just want to put a whole bunch of like unstructured

health stuff in there.

And then I said, well, how could I make this healthier?

How could I make this healthier?

And it was like, Oh, do I get rid of egg whites?

I was like, okay, no, I don't want it.

It's taking healthier as less calories.

I said, make it more nutritious.

Yeah.

And then it started giving me other ideas for superfoods or whatever to

put it into it.

So anyway, uh, it's kind of like everything you need.

It's like a Swiss army knife.

It feels pretty game-changing.

I have to say.

Yeah.

Well, the one thing, you know, that that's approaching and the way you're

using it is, um, I think like Larry Page always had this thing.

Like, you know, when he talked about say like car purchasing sites is

something you're going to use every like five years or seven years, but like,

yeah, he's always focused on something you use every day or multiple

times a day, right?

You can, you know, think about their core product and stuff

that they bought like YouTube.

And so it's approaching that, right?

With you.

It's starting to approach it.

Yes.

Yeah.

Cause I do take pictures of things and then I was just thinking, huh, why

isn't my phone, my photos app just using AI all the time?

And I guess Google photos kind of does like the face thing, but

it's not doing anything too crazy.

Um, but I feel like chat GPT-4 is ahead of everybody.

And with this multimodal thing, I think they took a step function up.

You agree?

Oh, they have.

And, you know, they launched something this weekend.

You know, I don't have access to it yet.

I was checking this morning where they've also made it that you can give

it like any type of document.

And so what you can do is like a PDF and so you can use that for, um, you

know, a lot of different use cases, which I think is it's going to take

it to another level, like, you know, from your work scenario where you've

got a document, you want a quick summary of it.

And, um, and look, I think it's going to become like a multiple use a day,

like a Google, Google search used to be.

And I think that's the threat for, for the folks at Google right now, right?

Why they're working really fast.

They did a big additional investment in anthropic this last weekend as well.

So.

Wow.

It's a.

And then I noticed also chat.

JPD for is now up to date on the web.

They, they, you know, they, they were saying like, Oh, this is up to date as

of like 2021, 2022.

Yeah.

And I think now it's up to date as of like September of this year or

everyone's getting different results.

Like when I type it in, it was, I think it's saying something different

for every person.

So I don't know they haven't said an official message, but like next

Monday is a big day.

They have their first ever, like developer conference, developer day Sam

tweeted a developer day, and it's going to be big, you know, from what I'm

hearing to the folks that we're in touch with, they're like new APIs that

you're going to be incredible.

I think they're going to do a lot of stuff with agents.

It's going to be really cool.

Building startups is really hard and things can turn on a dime.

You can do everything right and you can still fail.

That's where arising ventures comes in.

Arising ventures is a holding company and they acquire tech startups that

are facing setbacks.

Let me tell you about one of arising ventures success stories.

Up Council is a leading online marketplace where companies can find and

hire legal help.

Up Council was facing legal challenges and a slowing growth curve.

As they were about to shut down, arising venture stepped in to help turn

things around.

Within a week, they crafted a deal that worked.

They acquired the assets, they took operational control and they did this

without transitioning any team members.

They got the business profitable and now Up Council is serving millions of

people a month.

Arising Ventures knows what founders care about because they aren't bankers.

They're actually tech founders themselves.

So here's your call to action.

Arising Ventures quickly got Up Council back on track and profitable

and they can do the same for your startup.

So go to arisingventures.com slash twist today to learn more and connect

with the team.

That's arisingventures.com slash twist.

All right.

So let's see some other.

Okay.

We kind of beat the chat GPT for into the ground.

Do you have anything that's not chat before?

Yeah.

Yeah.

Let's let's do that.

So I'm going to go.

It's like one of my longtime business partners, my co-founder of my previous

two businesses.

He's launched something which I think is really exciting and this kind of

shows like kind of a vertical use case, which you probably will never be,

you know, taken out by like chat GPT.

A lot of people talk about what happened this week and taking out simple

applications and this intersects with your old a little bit.

So I can't wait to get your feedback on this one.

So basically.

Okay.

Mantel.

This is called M-A-N-T-L-E.

Mantel.

M-A-N-T-L.

Yeah, exactly.

Company ownership from the future.

Yeah.

And so it's like a kind of like a redo on, you know, CARDA and CAPTABLE management,

everything that come with it and company management.

And so, you know, I've got access to their product here and what you're going to

see is like, you know, and you've done this a bunch of times now you have your

team do it.

And so if I go and I'm just going to drop in company docs.

So this is, I just kind of created a new company.

I'm going to do a live demo here and I'm going to drop in company docs.

And you can see here what it's going to do is start processing those documents.

And there's a whole bunch of docs.

They just gave me a folder full of things.

And so here, and it's like, it's a dummy companies.

We're not looking at anything specific, but there's the original certificate

of incorporation here is an amended one.

So I can say, oh, you know, add this year class, the comments would have been

the founders, here's the preferred shares.

You know, here's a looks like probably it was like stock option agreement

that occurred.

Amazing.

Here's some share purchase agreements.

And so I'm just going to quickly go through and just add all these things in.

And, you know, J.K.

You've done this a bunch of times and your team does this.

This is due diligence, right?

You and it's a document locker and usually it's a document locker, but

it's not intelligent.

You can do a search on it.

You can sort it by date modified, date added, but that's about it.

You're going to be hunting and packing in and out of documents.

And even creating it is work between your maybe one of your leads plus

your lawyers usually and someone there.

It's going to cost you a lot of money.

So I basically added all those things in and you can see here now I've got

a fully populated table with all the relevant information.

And like this was work for a real person with like, you know, like a

locker or like a lot from a couple hours.

And this just happened in, you know, whatever 30 seconds here for us.

And then, you know, beyond that, I can kind of click in.

I can say, okay, show me the commons and I can go in here and I can see, okay,

you know, Michelle, I'm the share to see apps.

I can click on this and then I can see the vesting schedule.

You know, all this stuff, which is so hard to kind of pull out of documents

and look diligence.

And I think you've come across a few different things in diligence in the

last couple of years.

You've mentioned in a couple of different places as well, right?

I mean, often the problem is somebody might change counsel or the same

firm might change lawyers.

They use a different document.

One document can flick with another.

Then two years later, you find out, oh, you know, we gave these eight

people this document.

These two people got the old template.

These eight people are vesting according to this schedule.

These two people are getting a different type of stock.

One person had one confidential agreement.

Then you have to go get everybody to sign them again.

You have to explain to your team members.

Hey, we're resigning.

We're changing this.

And then they got to get their lawyers involved.

This is incredible inefficiencies.

Yeah.

And so look, all of this work is kind of, and this is where AI is awesome

because what it does is, you know, they've used it to take it, process

the documents, look at it and, you know, obviously a lawyer would look

or someone or your team would look at those things and match it up

when you were going through that process.

Like to your point, can you do this better than chat?

You'd be before you could dump a bunch of documents to chat to you before,

but it's not going to become your single source of document libraries

with your cap table in it.

And you're not going to get all that functionality around it.

Yeah.

Exactly.

And then what you can do is, you know, they've kind of got further down

and like they can look at those documents and create like a burn down

chart of your, of your equity pools.

You can see.

Wow.

Let me just zoom out.

Here's when our equity is fully, yeah.

Exactly.

Like, you know, what is a, you know, how much available do you have

in the option pool?

How fast is it burning down?

And this is the power.

Like you've kind of merged these things together.

You know, before you'd have to get your lawyer and even I've done it

in my previous company where you have to get a lawyer to create

and burn down charts for you and understand where, you know, people

are in their days of work, weeks of back and forth.

All done in an instant.

Brilliant.

All done in an instant.

Yeah.

So this is going to be great.

I think it's actually going to be great for attorneys.

It's going to be great for society because now what's going to happen

is instead of getting a $20,000 bill for setting all this up,

you're going to get a $6,000 bill or $12,000 bill and you're going to be

able to read a book or whatever it winds up being.

You know, let's say it goes down by half.

Now you can really redeploy that money.

And you know, that's what happens with safes.

That was an interesting or convertible notes.

That started to save people a bunch of time.

That's what happened with cloud computing.

And then what happens is more people can spend their money and their time

and time equals money.

We all know that money equals time, you know, in startups.

Then you can deploy that towards your product or customer acquisition

or whatever else it is.

So this is going to just be great.

More startups will start.

More startups will get, you know, to product market fit.

And this is great for early stage investors.

Exactly.

And then, you know, you're also able to kind of look at a lot of, you know,

nuanced use cases like of start dates that are different than vesting dates.

And so they have all these nuances in here, which, you know, we won't go

into, but you know, reach out to these guys.

They've done an incredible job there where, and those things are very hard

to track because sometimes your start date doesn't match with your start

of your vesting date and you want to move those things around.

And so all of that can be fully handled and, you know, cleaning

that stuff up is, you know, also expensive.

Well, and then you can do all kinds of scenarios with this.

You could say, uh, scenario planning is something people don't do,

but this will let you do scenario planning.

So remember last week we had the fall on the Google Sheets.

I want to show you that.

I want to show you that.

Because you could say, Hey, what if we get acquired and it would give

you like 10 different scenarios of getting acquired?

Check this out.

Cash or stock, et cetera.

So they have this.

So this is the existing cap table here and I can upload a term sheet.

So I'm going to upload this series C term sheet.

And basically it's going to be.

Here you go pro forma.

Exactly.

And I can say, okay, well, this is what it looks like.

If I just take it at is a $2 million round valuation cap nine option pool.

So I said, okay, if we change this to say, you know, 12 million, right?

That's what it'll look like.

Zip, zip, zip, zip, zip.

And again, do this with lawyers.

I was just doing this last week with a company I'm on the board of where

we had to go back and forth with the lawyer and say, Oh, you know, we were

negotiating a term sheet.

And so now you can just do the scenario planning here and say, let's

expand the option pool by 5% and so you can see what happens here.

Yes.

And so, and to me, the interesting thing is to ask questions here.

So now you could say yes, if the company was sold for $100 million, $50 million

in stock, 25 million in cash and 25 million on earn out.

When would, how much money would each employee get on signing year one, year

two, year three and year four?

And it would literally just give you each employee's earn out.

Yeah.

And like that takes months.

That's like, this is why M&A doc can take cost a million dollars to

do an M&A transaction.

Yeah.

So that's the next thing on their roadmap, J.

Cal.

And one of the things is, you know, they want to see if we can get launched

to, you know, an insight.

Are they raising?

This is your friend's company?

Did you slice in a little bit of a slice?

A lot of my business partner, we can work something out.

Let's get something done.

All right, let's talk about it.

I mean, if he doesn't have any bond there, sometimes you want to have J.

Cal as your first investor.

That's my new thing is I want people to have to say, J.

Cal gave me my first check.

I'm making t-shirts.

You know, I did 60 of these 25k first checks in the past year.

60.

I saw that from founding university.

I saw what you also did for the Y Combinator folks as well.

Yeah.

Your team really did it there incredibly well.

Yeah.

Well, you know, it's basically 1% of people get into Y Combinator.

And so I said, listen, you know, all due respect to Y Combinator, the

truth is, because I do this for a living as well, we can't tell the

difference between the 1% and the 10th percentile.

Probably even the 20th percentile, if I'm being honest.

Which is to say 9 out of 10 or 19 out of 20 people are just as good as

the top one person or startup.

And because companies pivot, like I think Paul Graham said like 50% of

people in Y Combinator wind up pivoting their idea.

I don't know that it's as high as that.

I think it might be more like 30, like a major, like a completely different

idea.

It might be a third.

Yeah.

So if that's the case, then even the selection process is just based on

your, your character of the, and the composure of the team and how

they present in 10 minutes.

Do you agree a little bit as well?

Maybe on the margins, some pedigree.

And so, you know, this is where I just told everybody email YC at launch.co,

your application before you apply, after you apply, after you got rejected,

whatever.

And my team will take a look.

And so we're doing 60 of those meetings a week.

It's not easy.

We do like a quick introductory meeting.

I give this with mantle gets a B plus.

This is great out of the gate.

It's great.

Awesome.

Out of the gate.

That's my little and they need to add the AI features.

The room for improving features.

The question and answer.

It is doing air right now, but it is not giving me like a bunch of

suggestions as to like, it should be anticipating my next question.

So I like, and you have this in chat.

We're at the bottom.

It will have suggestions.

That's like, that's like the low level version of it.

The better version is I looked at your data and based on the data, here's

the questions you should be asking.

So it should say you should really be asking us about a refresh because

you need to refresh.

You're going to be 80% out or you should be asking us about non-compete

because you're non-compete is really, you know, non-standard.

It should be anticipating.

And then it would be like not just a low level attorney.

It would be the high level attorney.

So it's not the associate at the law firm.

It's the partner.

This is doing associate level work.

That's why it's a B plus.

When it starts doing partner level work, then it gets the end.

But really nice job.

What do you get?

What do you give it?

Uh, I give it an A.

I know how hard it is to get, you know, these AI tools to look at.

How do they improve?

How do they improve?

Well, I think they, I think a plus is what, you know, is to add

the stuff that you're talking about.

Little great information.

Okay.

Yeah.

You know, all right.

I think they've done a good job.

That's really, really exciting.

All right.

Way to go.

Mantle.

And, you know, I think every time a new platform shift like this happens,

every single company, whether it's Carter or Uber or Amazon or Google

or YouTube, the whole table gets thrown upside down and you have to

pick the table back up, put the chips on it, put the cards, shuffle them back

up and you got to start the game over again.

And sometimes people still have a big stack.

Sometimes other people lose their chips.

So this is the great like flipping of the table.

And that means an opportunity for startups.

If you run a business, you know that having reliable vendors is non-negotiable.

And whether you need office snacks, holiday gifts or wholesale ingredients,

you need to check out nuts.com.

That's it.

Nuts.com.

That's a crazy, amazing domain.

N-U-T-S.com and nuts.com is your one stop shop for the highest quality foods

for your business.

Again, they offer delicious office snacks, corporate gifts and wholesale

ingredients.

I got a gift pack.

I have been eating these beautiful roasted nuts and other amazing premium

products like chocolate covered sweets.

I love the trail mix, popcorn.

You know, I stopped eating the candy.

So I went for the dried fruit, but they also have wrapped candy as well.

And my favorite, jerky.

And of course, they have all the gluten free stuff or whatever dietary

option you're into over 50,000 companies choose nuts.com for their

business needs from offices to hotels to restaurants to retail stores.

Nuts.com has something for every business.

And so here is your call to action.

Nuts.com makes ordering for your business quick and easy.

And right now nuts.com is offering new business customers a free gift with

purchase and free shipping on any orders of $125 or more at nuts.com

slash twist.

Go check out all the delicious options at nuts.com slash twist and

you'll receive your free gift and free shipping when you spend 125 or more.

That's nuts.com slash twist.

Let's go to the next one here.

I'm going to cue this one.

I thought this was really, really cool.

And, you know, it took a little bit of work to set up.

And I think the demo might be a bit tough because it tries to use Zoom to do

the demo, but we're going to give it a try here.

So basically this is called talently.ai and it's an AI interviewer.

Your AI interviewer.

Okay.

This is breaking like seven HR rules already.

But okay.

Yeah.

And so what you do is you go through a little, you go through a little form

process and you put a job description in there.

And I've went and done all that.

And basically out spits like you're, and I did kind of, I've created one here

called, you know, summer intern software developer for definitive intelligence.

Yeah.

You did your legal document last week.

Okay.

Yeah.

Now you're doing your interviews.

Like I'm, you know, I call this AI all the way down, Jay Cal.

Your company.

All the way down.

If you're building a company now, you should try to be AI all the way down.

Yes.

Right.

From your account table to, you know, interviews to catering.

I don't know.

We've got to figure it out.

Yeah.

Yeah.

That right.

I mean, we'll see.

And so basically, so I went through the process and set it up.

And basically what this will do now is, and I think this may, this is where it may

not kind of work here.

But I'd say, let's say Jason and, you know, welcome to your interview.

Okay.

So it's like the form you fill out.

Yeah.

And so this is what you'd send out to your, you know, perspective person.

And, and, um, let's see, hopefully it, it, it, cause it tries to, cause it's meant

to be like a zoom interview with like a lot.

Ah, and I hope it doesn't like kind of bomb here.

All right.

So show me what happens when the person clicks on the link and does the interview.

Okay.

So, uh, you get to this little box here and I'm going to click start interview.

And you, you have, you have, it forces you to share your entire screen, because

it wants to make sure you're not doing anything else because this is a coding

interview as well.

No cheating.

Yes.

No cheating.

And, um, and so here it goes.

Jason, we're pleased to have you join us.

I hope you're having a good day so far.

Are you getting that?

I am.

I am talented AI and I will be conducting your interview today.

This interview will consist of two stages.

First, the technical interview and then the live coding interview where you will be

presented with new problem solving questions to solve in real time.

To start, can you give a brief introduction about yourself?

Oh, hi.

I'm Jason.

I am a sim like developer.

I am a 10 X coder.

Thank you.

I don't know if it's going to work through our zoom that way.

No, it's not.

All right.

We, we see what's happening here.

So it's going to walk you through the interview question by question.

Yeah.

And then, uh, it's now throwing you into, uh, a developer window.

Welcome to the coding portion of the interview.

Oh, God.

Please solve these questions to the best of your abilities.

So, so crushing.

You have 15 minutes for each questions.

I don't think it is.

Two problem solving questions.

This is just your time starts now.

Good luck.

Write a Python function to implement a simple.

We value your input as a human.

The function is to take in your best job coding.

So we can eliminate your job in the next 18 months.

We're not using your intelligence to train this model and remove your

necessity in the planet going forward.

Please type.

You did not obey.

You will now receive an electric shock.

Please put your hand on the keyboard to receive electric shock.

I mean, they need to use a better, uh, text to voice.

Right.

They've used like so, so crushing.

Yeah.

Right.

Um, also, and it's like super wordy.

They didn't make it personalized.

Like use a celebrity voice.

Have it be like, you know, Arnold or Obama or something.

But Jason, we live in this world now, right?

Where we have these job shortages.

This is a coding interview.

But like, if you tighten this up, yes, this is, this is one of the, the, this is

one of the risk, like the limiting factors in finding people, right?

As you can't interview people fast enough, can't interview enough people.

Yes.

And you do it across more folks.

And I think, I think in a world of a lot of remote work, people working

from different places, I mean, again, I legitimately am going to use

this product for us.

There was like a moment in the planet of the apes where like, you know,

how the original one or the, the remakes.

I think this is like one of, no, the original planet of the apes.

I think there was one where like it's reversed.

And now the apes humans are on planet earth and they're trying to

figure out how zero whoever is like super intelligent.

And they're like moving the cones around and they're like, do.

And she like looks at them like, are you guys dumb?

And she just goes, puts it together and like, oh my God, this like monkey

human can do that.

And she'd like, then all of a sudden starts talking.

Well, of course I can do it.

And they're like, oh my God, it speaks as well.

That's how I feel like this is like the super intelligence is pandering to us

mere, mere humans.

Oh my God.

Okay.

Now I need you to write a piece of code that allows people to write a

comment and retweet what's on the page.

You have 90 seconds and it's like, it's done at 17 times in the first second.

That it took you to figure it out.

Remember, we don't want you to use a co-pilot or cheat.

You have 59 seconds.

I mean, and this is going to apply like you're going to be at McDonald's

and it's going to be like, okay, you're now making a filet of fish.

Do you remember what to put on a filet of fish?

And you're going to be like, Tartar sauce fish.

Wrong.

We put the tartar sauce on the bun, not on the fish.

Please try again, human.

And it's literally going to do this for fast food.

This is our lives is going to be proving to the robot overlord that we deserve employment.

Oh my God.

It's so great.

I give this a hmm.

It's pretty darn good out of the gate.

It needs polish, obviously.

But this is going to work for a frontline interview.

Pretty darn good.

Yeah.

I'm giving it a B minus.

Yeah.

Giving it a B minus.

Very quickly could be a B plus.

Yeah.

Very quickly.

I give it a B.

I'm already using it.

And I just think like the same thing you said is that it needs all the way down better.

It needs to have a better, you know, text to voice.

Like they used like a, you know, maybe a cheaper one or I don't know.

I maybe have to upgrade to get a better one, but like it's super useful.

Doing interviews is hard work, filtering folks and giving more people a chance.

There's so many people in their world that are talented.

Sure.

I can't, I can't interview a thousand people, but I'd like to get a thousand people to apply.

This is a great way to do it.

I mean, I want to do this with like a simulation for flying planes, like literally just give

a bunch of 12 year olds access to flight simulator for you around the world and then

just be like, hey, fly the plane and then be like, okay, you just got a bird strike.

Oh, you got a second bird strike.

Oh, the engine's over.

Flip it over.

Drink like two screwdrivers and then flip the plane over like in flight.

You know, it's going to be awesome.

This and also like it's going to take out all bias, right?

Like if you're like a short fat dude or you're like a tall handsome dude or you're,

you know, whatever ethnicity, all that gets stripped out on this.

It's just going to be pure meritocracy and that's good for the world.

This is like blind auditions for an orchestra.

Remember the movie car or in blank, they put people behind a shade when they, you know,

are casting like a new cello player in order to not have biases.

And that is like a really good thing to do in life.

And this is going to be amazing.

Just let a thousand people, you know, a second take these tests and yeah, find them and sort

them.

It's great.

This is awesome.

I give it a B minus.

I think this is unlimited upside is my little quick, my little quick take my little hot

take unlimited upside here because you could do this for a lot of different things.

We do writing tests.

We do all kinds of tests.

So yeah, I really like it too.

I think like the interviewing problem is like the biggest one and this allows us to open

up the doors to many different folks and I give it a B.

Just, you know, we're in the same zone now.

We're tying it up in the coding interviews thing is really neat as well.

And like it only took me a few minutes to set up that job description.

Yep.

Also, we're using this the next time we have to hire a video.

Absolutely.

If this exists for video editing, this is the greatest filtering mechanism of all time.

Yes.

You're actually getting people to do the work.

I just did this.

You know, I just hired, I'll tell you the filtering mechanism I use.

We hire researchers to come to the company and sort through all these applications.

I'm talking about startup applications.

You got like 20,000 of them now.

Yep.

We actually had too many startup applications.

So it's really hard because I like everybody to get an answer from us.

And listen, we meet with 15% of those 20,000, right?

3,000 a year.

Wow.

Okay.

It's crazy.

Yeah.

But we have to say no to 85,000.

I want to say no gracefully.

I want them all in the database in case they start another company.

I want to track everything.

So I hire researchers out of school and then they become analysts and then become associates.

So it's like a career path, right?

Yeah.

If you want to break into a venture.

But the key thing is I want to see them on camera interviewing a founder because I

believe that with founders, you have to have a certain amount of enthusiasm,

engagement.

It's not easy for everybody.

Not everybody's personality is designed to be engaging on Zoom.

Yes.

Right?

It's like, how do you account for that?

No, I know it when I see it.

Yeah, yeah.

So I just said, you know what, they were thinking about hiring one of like five people.

I said, have pay them because I don't like people to do free work.

That pisses me off when people like ask people to do free work during the interview process.

I said, pay him a hundred bucks, tell him to do a 20 minute interview and then have

one of our team members present their startup.

So we have a team member who has got a side hustle, Prash.

She's got this like app for cold plunging shout out to Prash.

So Prash, it's like a, you know, it's like a side hustle.

It's like, you know, like a zero for cold plunging.

Yeah, zero for cold.

Like Zach, on your watch, right?

I could have used it today.

I was in the cold plunge today.

That's why my energy is so great.

If I do say so.

And so he pitched it to her and then they sent me the video and then we gave her the

script and we gave her what questions we like to ask.

And I thought, well, she needs more training.

Yeah.

But really good out of the gate.

And if you see somebody do the job, you know, if you get, and if you see somebody

to do the job, then you know they can do it.

And just this whole like performative, you know, a hiring process.

I don't like, I'm going to either hire people going forward, having them do some

project for us or having them be freelance for a couple of weeks, 20 hours a week

or something.

If that's possible, sometimes not possible.

They're leaving another job and they want to have job security or whatever.

I understand that.

So anyway, this affords you that, right?

And to watch people work in real time, like Nick needs to know if they're a video

editor, can they do it fast?

Like somebody could copy somebody else's reel and do something really sexy.

But if it took them two weeks to do 60 seconds, that's not good.

We need somebody to do 60 seconds in an hour and make a 60 second reel in an

hour or two.

That's really good.

Yeah.

And I like how it forces you to like share your screens.

You got to see them do it.

So they can also.

Yes.

And then once that group of people, once that group of people finishes their

application on the software, then you can filter for, okay, will they be a good

team member?

Can I see myself working with them?

Yes.

Not the other way around.

The most important part is the work.

Exactly.

The work is the most important thing.

Yes.

Of course.

Okay.

Here we go.

This is why when you go for a job to be a chef somewhere, they have you make

a French omelet because a French omelet is like, if you got that technique,

right?

Your technique for other things is going to be just fine because the technique,

if you have really refined technique for a French omelet, look it up.

Gordon Ramsay or a couple of other people make really great ones.

Listen, point blank.

If you care about privacy, you need to check out Brave.

That's the browser I use day in and day out on my mobile phone, on my desktop, Mac,

PC, all of it.

Brave's browser now has over 60 million users and they use it because it shields

you from all these trackers and ads and all the other creepy stuff that follows

you across the web.

And it's built on Chromium, which is the open source Chrome project.

So all of your favorite Chrome extensions, and I use a lot of them.

They work perfectly in Brave and it's three times faster than Chrome because Brave

doesn't slow you down with all those ads and trackers.

Of course, you can import your bookmarks, your passwords and settings from Chrome

with one click.

There's nothing for you to do.

You can just start protecting your privacy and it's amazing.

Your whole web experience speeds up 20, 30, 40 percent every time I use it.

It's so snappy.

Brave doesn't track your website visits.

They don't track your searches.

They don't track your clicks.

That's all your business.

Brave Search is a truly private and independent search engine.

So download Brave today at brave.com slash twist to browse faster, search privately

and so much more all in a single click.

Give me one more.

I'm so excited.

I got two choices.

I got a game and chat, GBT 20 questions.

Okay.

Interesting.

Or we've got a kind of like, you know, again in the AI all the way down, something

that helps you write descriptions and tags for your YouTube videos.

And I thought we could try that as well with a YouTube thing.

I like that.

I like that idea.

This is a really important thing because yes, you know, the you guys are working through it too.

Yes.

So let's see this.

Okay.

All right.

Never worry about video descriptions, community posts, video ideas again.

Great.

So it's Maurice AI or yeah.

Morris AI.

I don't know.

It's a guy.

Okay.

So you click on your videos.

Yep.

And basically you drop it in there.

You give it the, the, the URL.

I've done one here and I'll show you and it'll generate the titles, the descriptions, the

icons, everything else that's needed for it.

And then basically you just have to copy and paste it back and forth.

And so wait a second.

Isn't this happening inside of YouTube now, Nick?

Didn't you say that YouTube was doing some of this built into the interface?

But it's very intermittent.

It's not consistent.

It seems like they're doing some sort of a B test because randomly, like I'd say one

out of every five videos, it'll say, Hey, here's a way to punch up your title.

It's called like titles with AI or something inside of YouTube itself.

So they're definitely testing it.

All right.

Next time you have that, make a screenshot so we can see it.

So this is a little bit more holistic.

Give you a couple of different choices.

So what I would like to do, this is, you know, I would like to A, B test this multivariate

test it.

And so what would be nice is if I could take this, make ads out of it, promote the ads

and then see which one actually drives the most clicks, but you can't do that with YouTube.

You can't change.

You can change your thumbnail.

I guess you can change your title.

So I guess you could put it up.

Hmm.

You could put it up and then I guess test three different ones with like $500 in ads

against each one.

Nick, you think you could do that?

Yeah.

Yeah.

So maybe you could figure that out, but I would want that all to be automated.

But this is great.

I think this stuff is going to save people a ton of time.

Yeah.

So I just did like a quick one here, like a, you know, making a sugar cake and it gave me

the description, gave me all the tags.

And so look, I think if you're in this world now and, you know, we should give it a try

for, for at least these episodes or some of the other ones and you guys do a ton of these.

And so, and they just made it super simple.

They have like just a two page website.

You go here, you log in and basically you open the app and you give it your URL and

it just off to the races on its own.

And so I did one for, you know, one of the previous episodes that it's that.

And so just super simple, like the way they've done it.

It can also give you ideas and other things as well.

Community posts.

So everything around the YouTube ecosystem, simple.

It's so smart.

I really like it.

Yeah.

I mean, produce AI producer is clearly a huge win.

And we're just starting to see it with our production team here.

We don't quite trust the clipping software yet, but it might be able to eventually get

a 60, 70, 80% of the way there, maybe in the next year.

And then eventually maybe it will work.

I would like to just see the entire transcript.

This is what I was telling the podcast AI team to show me the entire transcript and

let me highlight it and hit publish or clip and then make a clip, but make it dynamic

and interesting, you know, and give you three versions of it.

I watch all three versions of it.

And then I pick the one I like best like three different styles beta testing with

this company storylines, which we'll do a demo for it probably in next week.

But we've done some of the tick tock posts with them, right, Nick?

Yeah.

Yeah, I give this one a C plus.

It's very basic.

It's basic.

It's not doing anything that wowed me.

I think all this stuff will be built into YouTube.

So I do think that this is not as they're going to have to really up their game.

And maybe we didn't see all the features, but they're going to either up to the

game, the stuff, the basic blocking and tackling of tagging, titles, description,

all that kind of stuff.

That's just going to be table stakes and there's no reason to do it.

So it needs to, we need a more unique approach here.

And so I think sometimes entrepreneurs go for the easy stuff.

And then you get run over.

You got to go through something a little bit more chunky here, a bigger problem.

And the bigger problem might be like, what should I talk about next on the pod?

You know, analyze the comments.

Who in my, who should I engage with in the comments, right?

Who, what are some topics that I should discuss on my podcast, right?

That's kind of like this production and there's producers production.

That's what this is.

This is like little checkboxes in production.

All very important.

World-class production can change the game.

As we see with this week and startups all in other pockets that have great

production value.

That's why they call a production value like it matters.

Then there's producing.

And that's where like ideas, creativity, you know, formats, you know, coming up

with unique things that are hooks, the community.

And I would like to see this kind of graduate into the producing, not

the production area.

I want to give them one credit.

So basically they do have a video ideas.

And so I just dropped in is like, you know, one of the features that were kind

of deeper in the product.

So you give it your URL and it gives you a video idea.

And I guess decoding Tesla's energy plant.

Wow.

With Elon Musk.

Oh, what a great idea.

Get Elon Musk.

That's a little obvious, right?

Okay.

Wow.

Well done.

Yeah.

Brilliant.

Great.

Wow.

Hey, Nick, let's do that.

Yeah.

Hey, Nick AI.

We might get, we might get some views.

Yeah.

I'll ask Larry Page and Sergey next.

Yeah.

Oh, wait.

No, here's a better idea.

Have Zuckerberg MMA fight Elon Musk in the Coliseum.

Oh, that'll get a lot of views.

It's like, we're going to need you to take it down a couple of notches here.

From, uh, yeah.

Oh, great.

Okay.

I had to see, I'm going to give him a C plus because they did this.

I'm going to give him a C plus.

Okay.

I'm, I'm giving him a little extra credit here.

Another amazing episode.

What was your rating here?

I'm going to, I'm going to line with you.

C plus, I think this is really cool.

The, you know, we've got to dive deeper into it.

Hopefully they reach out.

We learned more from it.

I think it's really interesting.

Keep grinding.

This is, you can do it more here.

It's really cool.

Well, and this is part of your journey as a founder.

You put something out there in the world.

People get use out of it and then, you know, you listen to your customers.

You, you tackle something even harder.

And so yeah, maybe this is helpful to people, but you know, it's got a shelf life

and it's probably going to get, it's probably roadkill.

As long as you know it's roadkill and it's going to be a feature inside of

YouTube very soon, um, then you're freed up to just not worry when that happens.

And then think about the bigger picture, which I like the ideas.

I think the ideation is truly unique.

I would stop this other nonsense and just focus on ideation.

Like if they, if it told me, here's what all the, here's all the techniques

that other, you know, folks are doing on YouTube that work that you're not doing.

That would be very interesting to me.

You know, that would be more interesting.

So I would put 95% of their effort into that 5% into this kind of stuff.

But what happens is this human nature, you know this, Sonny.

People will do what's like easy, right?

They want to get through their email inbox.

They want to get through their Slack.

They want to do their expenses.

They want to go speak at a conference, but talking to customers, doing sales,

hiring, those are hard, high level, difficult things.

So people shy away from them.

And so I'm constantly in my life trying to get the people on my teams and in

startups that we invest in to not do the busy work, but to do the high order

work, to do the real game changing work, to do the difficult stuff.

And you know, listen, you got to do your chores, obviously.

So you got to do your chores, but you also got to do the groundbreaking stuff.

And so here, this was a lot of chores, not a lot of groundbreaking.

I'd like to see it do some groundbreaking as opposed to, you know, the one that

was Mantle, which was like, that was doing like, that was doing like the hard

high order bit.

Yeah.

And like, so it was like somewhere in the middle between chores and the high

order bit, but moving towards the high order bit.

All right.

Another amazing show.

What do you think?

We should do this this week in startups AI demo day.

Just have like 50 people come and demo.

We have to, right?

This is so exciting.

Yes.

All right.

So we have to find a location like 300 or 400 seats, no more than that.

We'll invite 50 people to come.

Yeah.

25 demos a day.

Is it even doable with everything?

Sure.

There's plenty.

I mean, all we need is a theater, just a theater and you know, some theater

we can rent for like 10K or something.

And we could do it.

We could also do it in the peninsula.

We could do it like in Redwood City or something.

You just get the same number of people.

Yeah.

So we can find a theater that holds 300 people.

Nice movie theater, nice theater or something like that with nice seating.

Put a nice lunch in in the morning.

We have 20 demos in the afternoon.

You have 20 demos.

That's it.

No.

Uh, yeah.

If we give everybody 10 minutes, five an hour.

Yeah.

You could do 20 in the morning, 20 in the afternoon and then have a big

long networking lunch done.

Yes.

This week is sort of AI demo day coming soon.

We just got to find a date and work backwards.

We'll see you all next time on this week.

All right.

Bye-bye.

Yeah.

Machine-generated transcript that may contain inaccuracies.