hi dear friends,

Here’s a look back at how our AI headshot app the Multiverse AI grew to its first $75k USD in revenue.

Here’s a timeline for you:

• April 2023: launched the Multiverse AI and quit my full-time job

• January 2024: $20k USD revenue (9 months after launch(

• June 2024: $30k USD revenue (14 months after launch)

We don’t have any investors, we are fully bootstrapped. How is that possible? Read here 👇

PS I first wrote this post in February 2024. I’m posting this now on the newsletter since I’ve looked back on it a lot, thinking about our past growth and how we can accelerate that further.

Stay hydrated and eat your vegetables.

Tanya

____

We launched the Multiverse AI late April, 2023. I’d just left my corporate job in marketing at OnePlus and OPPO. The world was abuzz with generative AI and I’d been playing around with Midjourney, ChatGPT, Stable Diffusion. 

Motivated by the bootstrap community, Anika, Zeeshaan and I got together and decided to build an AI SaaS headshot generator platform. Bootstrapping in the startup context refers to the process of launching and growing a business without external help or capital

It’s been so so many long nights (and days) grinding away. Days working while having a fulltime job, nights working on our project. 

It is HARD. It was a bootstrap from the ground up, zero marketing budget, zero investments, zero money raised. 

We launched late April. 

So... How did the first nine months go? 

Since April 2023:

• The Multiverse AI was recognized by and featured in Forbes, Superhuman and TalkAndroid 🥳

• We got to 75,000 USD total revenue in 9 months ⚠️ and closed a record month in January with 20,000 USD revenue ↑

• Shipped a TON of features, including our flagship b2c product, The Multiverse AI for Teams, many iterations to improve our accuracy, and developed more styles in partnership with HR and sales teams

• We've welcomed our first large teams, serving our B2B platform to teams of +200 members, demonstrating the scalability and appeal of our AI platform

• Forged meaningful partnerships that have expanded our reach and impact with organizations like Unite.ai, GeeksforGeeks, and many more

• We’ve delivered AI headshots to professionals from dream companies like Google, Walmart, McKinsey, Deloitte and more 👩💻

• Our product-market fit has been validated by overwhelming positive feedback and high ratings on platforms like Product Hunt (4.9/5 ⭐⭐⭐⭐⭐)

The Multiverse AI featured on Forbes (organic!)

Why share?

1. I’ve seen a lot of people in the bootstrap community share numbers, and it’s been incredibly helpful. Everyone’s journey is different. Ours isn’t the best, isn’t the worst. We’re getting there and learning every day. Sharing is our way of contributing to the startup community and I hope this helps someone out there - I know I’ve benefitted a tremendous amount from people I’ve never met sharing information I never even knew I needed. Through online sharing, I've connected with so many people who have been helpful and hope to pay that forward. If you think I could be helpful to you in any way, my DMs are open.

2. Another thing that’s so much more challenging than I thought it would be: highlighting growth. I wish I’d written something six months ago. Now, I look back at that period and I think “my god I was an idiot, I should have done x and y instead of z.” That’s growth. I hope I can keep looking back to the past and think my god, I was an idiot”. 

One of our many product roadmap session. As a remote team, we used miro for this one

How did we grow to +3000 paid users within 9 months?

1. Customer driven features! If your customer is asking for a feature, get on it and get on it quickly. Zeeshaan Lakdawala is our rockstar developer and is fantastic at swiftly implementing changes that our clients ask for. This also works well because you know it will benefit the people asking about it. That matters! It reminds me of my time at OnePlus where one of our key tenets was to stay in touch with the OnePlus community. Your users know what they want and they will ask for it. Listen!

2. Team-building is KEY. You are nothing without your team. We wouldn’t be where we are if we hadn’t paid attention to that. We are lucky to have expertise in fields with very little overlap: marketing, UX and software engineering. I’d recommend anyone who is building a company and looking for co-founders to start there.

3. Ship quick. Ship ship ship. It can’t be stressed enough how important it is to just get 👏 things 👏 out 👏 there 👏. Your first version is your first version. You can always iterate, change, upgrade, build more. But get that first version out there as soon as you can.

4. Pareto prioritizing. We’re only three people. Everything takes time. So much time. And there is so little of it. The Pareto principle states that for many outcomes, roughly 80% of consequences, come from 20% of causes. Focus on getting important things done and don’t get stuck in the details. My co-founder Anika Lakdawala manages to combine a day job, with raising a baby, with her work at Multiverse (with staying alive as a human being who needs to eat and drink and exercise). It is no small feat and on a base level, I have no clue how she does it. But I will bet it has something to do with her killer ability to prioritize.

5. If you build it, you need to market it. Building something only takes you so far. Educating people about your product, meeting them at their knowledge level and communicating your value proposition is key. We’re building a high-tech platform that implements AI to get people a great professional photo. I can talk about what the Multiverse AI is for days on end. But that doesn’t work. Ultimately, what is your added value? Focus on marketing that. Product market fit gets you far. Marketing gets you further.

Foldable beds are not uncommon at Chinese tech companies, where 996 culture and long working hours are normal. We'd all regularly stay in the office until 10pm. Here I am lying down on one while planning yet another smartphone launch.

In February 2022, I was working long (long) hours in China at a tech company. In February 2023, I had just moved to Dusseldorf as a PR manager. February 2024 is looking very different.

We’d launched Multiverse in April 2023, thinking that with a lot of luck and hard work, we could hit our goal of reaching 10,000 USD by the end of the year. We more than 7xed that. Our revenue growth month over month is at 45%. Quarter over quarter it’s 191%. Hard work pays off!

Looking ahead, our vision is clear: to establish The Multiverse AI as the premier AI-powered headshot platform for professionals, with a strong focus on B2B solutions.

happy clients = happy multiverse

Reflecting back, I’m filled with gratitude for the support and feedback from our users, partners, and the broader community.

We’re sleepy but hungry and ambitious and ready for more. We're thrilled to continue evolving, innovating, and serving our growing community of professionals.

In the next couple of weeks, we have some really exciting features that are coming out too. Stay tuned! 

What entirely AI-generated team galleries look like with The Multiverse AI. Note the consistent backgrounds (outdoor on the left, white background on the right), wardrobe, meticulous hair.

I want to also take a moment and give a big thanks to our early supporters.

Some of you have helped us with product feedback, word of mouth sharing, business advice, emotional support. Nothing exists in a vacuum and our business sure as hell doesnt. The little comments you've made along the way or sharing of our product as we're doing another 12 hour work day, it really helps. I appreciate you. Thank you ⭐♥️

_________________________
PS I first wrote and published this in February 2024.

Get your professional headshots here: https://www.themultiverse.ai/ 

Here is mine btw

Hello my dear friends, 

Last week, this happened. 👇

yes that’s me talking to digital trends about a capybara affirmations

How did that happen?

I will tell you.

We will cover:

• 🎉 how I build my first AI app and website as a marketer

• 📰 how I got to talk about capybaras for a respected media like Digital Trends (4 million followers)

• 💶 how and how much I monetized capybaras

ive been told to include more photos of things and myself for conversion

___________________________

A while ago - I was having a really hard time.

How come? 

-> The Multiverse AI has been doing well, we happily reached $30k USD revenue for June 2024 🥳

In July 2024, thanks in part to some amazing affiliate partners like Artturi Jalli, Andrew Best and others, we’re on track to beat that (not by much but fingers crossed). 

But still. I’ve always been fascinated with some of the online communities on the internet.

That sounds silly but hear me out. 

Part 1: build in public

On Twitter (or X 🙄), there’s this awesome community of people who operate under #buildinpublic. 

It’s so cool. 

They tweet about their startups, how it’s doing, website metrics, conversion rates, what they’re doing to improve all of those things. 

And I kinda love it.

It’s so fun to get a little peek behind what people work on, how they are actually growing their business. 

Some of these people include Pieter Levels or Marc Lou or Danny Postma

2/3 of these btw are Dutch and 1/3 is French, which for me (Belgian) I find very cool. 

On the internet, by rule of numbers, people tend to be American. So to find a group of successful solopreneurs who are European is like 🫶😍🥳👯

Here we have a conundrum... A conflict if you will. What is causing this?

Part 2: suffering and salvation

There is a dissonance between my desired and my current situation: 

please read "i am a marketer" in the meme voice of i am just a girl

I am a marketer. I can tell you everything about keyword difficulty, search volume, PPC, emails that convert, landing page best practices or winning assets. 

But

I don’t know how to build things, or websites, or apps, or anything like that. 

I see a block of code and I feel like..

Is this the matrix?

Am I Neo?

I shut my computer in fear and run away. 

I thought I had to overcome that fear and I felt I had to enter the matrix and it was paralyzing.

This is what I think coding is

Then again, I’d like to resolve the conflict between situation a and situation b. 

-> Enter the glorious marvel that is the internet. Where we can truly learn anything.

And so I set to learning tools like Webflow or LaunchLemonade🍋  , Pickaxe, Framer , CARRD, Inc. Figma FlutterFlow Glide Adalo and so many more.

Standing on the shoulders of awesome founders like Ian Eck Cien SolonKatt Risen

And consuming hours (so many hours) of content where awesome creators like Adrian Ching , Mizko , Ryan Hayward 👋 and Mohamed Ghaith explain (often meticulously) how to make things whilst NOT being a developer. 

So I stand (precariously) on the shoulders of giants, which allows me to do this:

dodging code bullets

It was SO fun. And SO hard.  

My mental monologue at this point: 

• 🥳 I will be a marketer who also knows how to build things 

• 😭 I suck at everything wtf are divs

• 🥳 This is going so well, I made a button! 

• 😭 Wtf are HTTP status  

• 😭 This tutorial is gaslighting me, there is no code override section

• 😭 The mobile website navbar keeps disappearing i will never learn

A bit of a rollercoaster 🎢 until I realized 

everything is gonna be alright. I don’t have to know everything. I can take things step by step. I am uncomfortable because I am learning new things. I am fumbling but I will give myself grace. That’s okay. 

did you just judge yourself for fumbling as you learn and try something new? damn give yourself some grace

A bit of time passes, I continue learning and fumbling and learning and fumbling some more.

I decide to build something to help give me that grace. 

part 3: capybaras and science

What do I know that I can build? Turns out I have a very particular set of skills...

a very particular set of skills

Which brings me to the capybara.

🛠️ Skill: capybara content

I love capybaras content.

And I'm not the only one.

Capybaras are very popular. Known for being "very chill", for being the friend of everyone (including alligators) and they love relaxation.

Specifically, as aquatic animals, they love sitting in Japanese onsen baths filled with oranges 🍊

the perfect image doesn't exi-

What's more - IG is flooded with capybara content. Accounts on Instagram like capybara_worlds have +750k followers.

Kpop star Jennie who's huge (85 million followers huge) has a sunglass collaboration themed around, you guessed it, capybaras.

jennie and me have SO much in common

🛠️ Skill: affirmations

Plus I love affirmations. Affirmations help me get through things. They "arm" me with positivity and help me attack or defend against whatever comes my way.

That's scientifically backed too, by a bunch of studies like the ones here: https://capybaraaffirmations.ai/blog/science-behind-affirmations

• Cascio, C. N., O'Donnell, M. B., Tinney, F. J., Lieberman, M. D., Taylor, S. E., Strecher, V. J., & Falk, E. B. (2016). Self-affirmation activates brain systems associated with self-related processing and reward and is reinforced by future orientation. Social Cognitive and Affective Neuroscience, 11(4), 621-629. https://doi.org/10.1093/scan/nsv136.

• Falk, E. B., O'Donnell, M. B., Cascio, C. N., Tinney, F., Kang, Y., Lieberman, M. D., Taylor, S. E., An, L., Resnicow, K., & Strecher, V. J. (2015). Self-affirmation alters the brain’s response to health messages and subsequent behavior change. Proceedings of the National Academy of Sciences of the United States of America, 112(7), 1977-1982. https://doi.org/10.1073/pnas.1500247112.

• Albalooshi, S., Moeini-Jazani, M., & Warlop, L. (2019). Reinstating the resourceful self: When and how self-affirmations improve executive performance of the powerless. Personality and Social Psychology Bulletin, 46(2), 284-297. https://doi.org/10.1177/0146167219853840.a capybara affirmation

and what's more - personalized affirmations work better.

• Cohen, G. L., & Sherman, D. K. (2014). The psychology of change: Self-affirmation and social psychological intervention. Annual Review of Psychology, 65, 333-371. https://doi.org/10.1146/annurev-psych-010213-115137.

i think intuitively i kind of knew that because when i was fumbling and suffering and dodging code-bullets, these were my personalized affirmations:

🛠️ Skill: AI

and of course i co-built and marketed The Multiverse AI.

That brings me again to my very particular set of skills ...

So Capybara Affirmations AI is born. Capybara Affirmations are images with capybaras and an affirmation, made with AI.

You can check it out here: https://capybaraaffirmations.ai/

And here is a capybara affirmation: (here are a lot more)

a capybara affirmation

It took me a couple of weeks to build the whole thing.

In this time, I'm of course still working on marketing for multiverse, and all the other time i spend thinking, dreaming and looking at capybara affirmations.

part 4: launch

hello are you still with me? we're like 1000 words into this story and prob 90% of you have churned (stopped) by now.

if you're still reading this - you are doing great 👇

a cute capybara affirmation for you the reader

It's time. it's july 15.

i have a website, i have an app. it's all plugged into stripe, we're ready.

i launch by tweeting about it

at 815 followers, i don't have a particularly high audience number on twitter (815. you can follow me here)

but the right people find me, like Rita Liao and Prakhar Khanna

and also ... a journalist from digital trends (the awesome Nadeem Sarwar )

cool cool

ok break.

intermezzo on digital trends.

digital trends is a HUGE deal (to me)

why?

-> from 2020-2022 i worked in tech PR & marketing. in that industry, you're constantly trying to get latest releases or features in top tier tech media. think cnet, the verge, tom's guide, wired, forbes, pcmag. and DIGITAL TRENDS (!)

a huge part of our job was to get into placement (coverage) in top tier media like that. so i know how much effort that can cost.

here's a fun screenshot of 2021 tanya telling a coworker (at oneplus) about which are "top tier media" 👇

so this ... is cool.

tears

part 5: metrics of mugs

my marketing brain has had to learn some tech cost things too:

• ai server costs are significant. for every capybara affirmation created, it costs $$$.

• hosting a website costs $ too

so everything costs money. so i need to make money. so it is monetized.

-> for a breakdown of the whole profit & loss, you can go here and read more about exactly how much i spent & made on capy affirmations so far**

here was the business model until then 👇

how capy affirmations works: you can make up to 3 free capy affirmations. then after, you pay $5/m to create more.

capybara affirmations app

ok so p&l notwithstanding and metrics aside...

i love capybara affirmations.

i love that people make capybara affirmations using the https://app.capybaraaffirmations.ai/

i love building something that shows how ai can be positive and bring joy.

i love that i actually built something (no matter how silly it is)

i love that capybara affirmations backend is full of

• ppl encouraging their friends 'you are so awesome and work so hard' 'im keebo and today will be a great day', and about 1000 other phrases (we're at a little over 2000 capybara affirmations generated iirc)

• ppl making cute birthday greetings

• ppl making cute affirmations for themselves

• or for their friends

• people sparking joy

it fills my heart with joy to think that someone in this world is getting their first capybara affirmation on whatsapp or email or anything. what a nice thing

-> capybara affirmations ai is a mental health app that scales positivity. with ai. and capybaras.

one of my paid users is a guy who created 12 capy affirmations. each one for a different name or friend (😭 my heart).

ok enough of the warm fuzzy stuff.

back to the numbers.

this is colder territory

traffic: pageviews 8.1k, nice. uniques 6.1k, also nice.

conversion to paid:

• 6.1k uniques (thanks mostly to shoutouts from marginal revolution, digital trends, a

• -> +2000 users

• -> +500 members (they sign up)

• -> 4 paid subscribers (one of which my mother bless her heart)

• conversion rate: 0.06% 😭

sad

read: wasn't converting well.

ps conversion means a user goes from top of funnel (awareness: user is aware) to bottom funnel (usually purchase: user pays).

if you're reading this, you are now aware of capybara affirmations. you are top of funnel for capybara affirmations

i am biased because The Multiverse AI is

• ✅ pay-to-play (no free tier)

• avg cross-source/medium conversion rate is 9% -> 🙏 blessed 🙏

comparatively Capybara Affirmations AI:

• ✅ cute and positive

• avg conversion -> ❌ not blessed ❌

so i had all this traffic. but very few converting

meanwhile, ai server costs are ramping up...

felt like people LIKED capybara affirmations, but maybe more to look at than to create themselves.

i pivoted.

i launched july 15.

-> on july 22, i implemented a shop that sells mugs: https://shop.capybaraaffirmations.ai/

my mug with a mug

this is relatively painless because of services like sellfy and red bubble that allow you to upload prints. they handle fullfillment, printing, shipping, everything.

honestly not bad - i sold 5 mugs so far (total)

who buys mugs? good question

• i bought one

• 3 friends & family, one of which, of course, my mother (bless her heart)

• a stranger (!) -> this is a win

the most 'popular' mugs are the ones that say 'i do not have impostor syndrome' and 'i am organized and efficient;'

here’s the tldr of monetization:

Twitter tweet

what can we do? what do you think?

i hope that capybara affirmations will be able to sustain itself (financially) over the next months so i can keep the website alive.

my take on options:

• 1. create an app for capybara affirmations that sends out push notifications w/ affirmations (how??? no clue but maybe i can learn? hire someone if its hard? feels $)

• 2. go hard on mug sales (do people want to look at capybaras and affirmations while having coffee? do they not? do i want to push selling mugs? we will find out)

• 3. try to sell capybara affirmations to someone who can and wants to take it further? (a wellbeing company? mental health?)

what do you think, dear reader?

do you have any ideas for Capybara Affirmations AI?

if yes please comment or DM me - maybe we can brain crowdsource our way to more capy affirmations in this world

#buildinpublic 😎

ps you see isnt it fun to see startup metrics? there's something here

so what's next?

the future of capybara affirmations... i'm not sure what will happen. can it sustain itself? will i learn mobile development? what's the demand for capy affirmations like?

i don't know.

what i do know

-> i will keep learning

-> i will keep failing and fumbling

-> i want to build more ☺

more apps that highlight the value of AI.

maybe more cute apps like Capybara Affirmations AI

or more serious apps like The Multiverse AI.

i dont know right now. but i will figure it out 🍊

-> also a big shoutout to all the early adopters & ppl i continuously and gratefully get to ask for advice and thoughts or who just sometimes like my things, i appreciate you so much 💞

who am i?

im tanya, i made and market cool ai apps like The Multiverse AI and Capybara Affirmations AI. im also a forbes cyprus 20 women in tech and an authority mag 'women leading the ai industry'. i do some public speaking and post on linkedin sometimes

-> read my last post where i talk about building the multiverse ai: How did The Multiverse AI get to 75,000 USD revenue on $0 advertising spend? 

-> follow me on LinkedIn for more stuff on ai and learning and fumbling :)

-> make a capybara affirmation here and affirm yourself with a mug if you want here.

/

Happy new year my beautiful electric dreamers,

2023 was a big year. I built a thriving business, moved to Barcelona and picked up some other things along the way:

1. The days are long and the years are short (do stuff now)

2. Embrace failure (be cringe)

3. Trust your gut (nobody’s gonna know)

Read electric dreams to stay ahead on the latest in tech and AI

8 things to know

1. News station to use AI avatars instead of human anchors. From Channel1, Expect to see more of this

2. Krea AI is my latest favorite free image generation tool. It adjusts in real-time and allows for composition changes. Remarkable

3. After a whole year of testing AI tools, I find that I can do almost anything with ChatGPT Plus. Yes there are some other tools I use a lot. But in the end, the ChatGPT subscription is 110% worth its value when used correctly

4. AI livestreamers are becoming common in China. “During Singles’ Day this year, JD.com’s [similar to Amazon] virtual influencers, called Yanxi virtual anchors, were active in over 4,000 live broadcast rooms”

5. New York Times’ profile on Jensen Huang, the CEO of Nvidia. Nvidia matters because they basically have a monopoly on all AI chips. Nvidia's head of marketing about the GPU “We invented the category so we could be the leader in it”. From the New York Times

6. It’s so real. Hello Kitty invasion of the US / Ethan Mollick, Twitter

Twitter tweet

7. Doing business in Japan. Your company will provide structure and purpose for your life. Great article on Japanese tech culture by Kalzumeus

8. Landing pages explained. Which landing pages are great and why

Stay hydrated, eat vegetables.
Tanya

PS we launched a bunch of new features for Multiverse (my AI headshot startup), including more advanced details and lighting for all headshots, and a feature that lets remote teams create a team gallery. Email me back if you’re interested in learning more.

Connect / follow here 

• LinkedIn / I’ll be sharing more AI tool demos on here in the coming year

• Instagram / My AI visual journal where I occasionally post image generation experiments @electricdreams777

• Hit reply on this email. Say hi or hello again. If I dont know who you are, link your LinkedIn or something. I’m keen to speak to more ~online strangers~ in 2024.

Stay hydrated, eat vegatables

Tanya

Reading time: 3 minutes

scroll down for the prompt

Hello hello hello! I’m in Milan for another week and in Barcelona afterwards. Hit me up if you are around! Subscribers might get access to an exclusive cerveza.

This week:

1. Adobe Photoshop implements AI. What’s next? 🤖

2. An Otter meeting assistant 🦦

3. Why AI can’t (& won’t soon) get eyes right (｡◕‿◕｡)

4. ChatGPT plugins that are actually useful? 🤔

5. Why LUCK is so important in image generation 🧸

auto expand with Photoshop’s “generative fill”

1. Adobe Photoshop implements generative AI tools. What’s next? 🤖

• Adobe, the legacy birthplace of fake imagery, implemented a tool that can add or remove objectives, change backgrounds with contextual accuracy and more.

• Canva also has these if you want to go and try it for free there.

• Image editing with AI is still magical the first time you try it, and I highly recommend you do.

• All this being said, generative AI is basically becoming mainstream.

• Not because almost every one (Microsoft, Google, Snap (bad rollout), TikTok (soon), and more have their own application. It’s because people’s parents are doing this. It’s within reach for everyone. The Facebook moment is kicking in. We’re reaching peak mainstream.

• What’s next?

• Think of it as the Spotify moment. Spotify was once the hot new tech kid on the block with its ✨ai-powered recommendations✨. Now, we barely think twice. It’s embedded. Nobody’s freaking out about radio losing their job. We barely remember a time before algorithmic music streaming. Spotify’s weekly recommendations still feel like magic to me, and that’s not entirely unlike most generative AI tools.

• A lot of the tools seem somewhat similar at the moment. We’re converging. Partially because many tools are based on ChatGPT (chatbot) or Stable Diffusion (open source image generation).

• Interfaces - this is one other avenue for innovation. Currently, we’re communicating via chatbot. But just like how the smartphone brought internet into our pockets, I’m curious what our next form of engagement with AI will look like. Perhaps videochat? Actually, that brings me to the next point.

Otter’s transcriptions

2. Otter.ai - meeting & transcription assistant 🦦

• Otter joins your Google Meets, Zooms to take meeting notes. At the end of the meeting, you’ll be sent the transcript including screenshots of presentations and more.

• I typically take pretty intensive meeting notes, breaking down from takeaways to actionables - but I’ve been using Otter for some work meetings and it’s been wonderful. I still take my own notes but much less.

• It feels different to have an OtterPilot in the room. It’s literally like another person joining into your Zoom call and automating your common intern duties for you: taking notes.

• It’s shocking how quickly you get used to always having Otter around. Whenever we’re not sure of what someone said last meeting, or a decision that was taken, we can just pop into the Otter notes and double check. Makes for a great workflow and second brain mindset.

• Pricing: free tier with up to 300 monthly transcription minutes.

• For $8/m, Otter can even join virtual meetings when you are double-booked. And then still send you transcripts. Kinda genius as it helps to free you up from meetings, without having to miss what was said. It’s like your very own meeting/podcast PA.

• Verdict: ⭐⭐⭐⭐⭐ that’s a five out of five

3. Why can’t AI get eyes right?

• This very commonly asked question betrays a wrong base assumption. You are assuming AI knows what “eyes” are.

• It doesn’t. Actually, an image generator AI like Midjourney doesn’t know what ANYTHING is. It doesn’t have the faintest idea. It cannot reason or think, it can only sort-of mimic what it’s been told. It’s a savant toddler. Read more about how that works here.

• So on the topic of eyes. On photos, eyes don’t always appear perfectly in the middle. In a profile photo, it looks like your pupils are glued to the side of your eye.

• That confuses the AI. It doesn’t know that eyes are only glued to the side when you’re capturing someone in profile.

• It doesn’t “know” when someone’s looking left or right, because data isn’t usually labeled so meticulously.

• Most training data descriptions will simply say something like “picture of girl with blue eyes”. So the AI mashes up the billions of images (including eyes that are glued to the side, like in a profile posture) that it’s gotten, and then spits it out. Result: eye mangle!

• TLDR; AI doesn’t know what anything looks like and mixes all possible eye positions up.

4. Fav ChatGPT plug-ins

• ⚠️ Plus users only. These are only available for ChatGPT plus users ($20 USD / month). My suggestion: try it for a month, and see if the time saved is worth it to you. I can say with a resounding YES that it’s worth it

• However, the browsing functionality is excellent. Turn it on by going Settings > Beta features > Browse with Bing + Plugins. To use it, hover over GPT-4 and click on “Browse with Bing”.

• What do I use the browser for?

  • When I’m traveling, I like to give ChatGPT my approx area, fav things to do, and have it plan out weekend trips for me. It takes into account distance, preferences, and can even add in the Google map links to all the locations if you ask it nicely.

  • Prompt: I am in [PLACE] at [TIME]. I like [INTERESTS]. Please create a plan for me this weekend. Give me different options.

• Let’s be real though - a lot of the plug-ins so far I’d rather... plug... out? They’ve been subpar. Here are the exceptions:

  • Show me: this tools lets allows you to create mindmaps, directly in ChatGPT. I’m very much a visual learner and this has been great

  • Wolfram: helpful for creating graphs and charts.

5. Why LUCK is so important in image generation

• For those (woefully) unfamiliar with image generation tools: people don’t talk enough about the factor of LUCK that is required for image generation.

• In Midjourney, every prompt generates 4 images. Then if I’d click regenerate, it would generate another set of four different images.

• One prompt can generate millions of different images. Whether or not you happen to find one you love in a first or second batch instead of a tenth or eleventh, is a function of luck.

• Skill comes into play too. That’s because you can fine-tune and narrow prompts down quite a bit. But the skill part has been overplayed (in my opinion), and the luck aspect is extremely underrated. 

• For example, out of the above four - I was looking for the “atmosphere” of the third photo. It took about 10 regenerations to get there.

• The prompt I used here:
  joonay the gyoza girl, in the style of mixes realistic and fantastical elements, matte photo, toycore, chinese iconography, berrypunk, colorful cartoon, realistic sculptures --s 750 - @electricdreamer

• You can try it for yourself! I’d be curious to see your outcomes (DM/mention me if you like @twizzles7 or LinkedIn)

• You see words like “toycore” and “berrypunk”. These words don’t actually exist. We’ll delve into that next time.

For now - stay hydrated, eat vegetables. I wish you nice human and non-robot things like sunny skies and solid sleep.

Cheers,
Tanya

PS.: I’m switching over to a different posting schedule so I can prioritize quality > quantity. If you enjoyed this, please share on your social platforms or forward to your friends as it’ll help other people find the newsletter. It’s also very encouraging for me to keep writing and see the subscription thing tick up.

Previous editions:

• Who will win in AI-generated music? 👨‍🎤 

• Why Can't AI Get Hands Right? 🙋‍♀️

• Beginner’s Guide to Generative AI ⚡

• AI agents and game designers losing jobs to AI 🎮

• How to use ChatGPT in Excel 💻

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

⚡ AI rabbit holes: therapy, hacking, nightmares, captchas and more

Reading time: 3 minutes

Hello friends. I’ve been commissioned a long report about how consumer brands can leverage AI. It’ll be available in a couple of weeks, after I pour more of my heart and soul into it. If you are or know anyone brand-side who’d be interested in answering some questions, would love to connect. So far, I’ve got people from Coca Cola, Nike and more.

PS This is the first larger project coming out of electric dreams, so I’m very excited. I’m also working on tailored trainings and more so - cool things coming.

This week, I’d like to encourage you to dive deep into a couple of AI rabbit holes.

Rabbit hole starting points:

1. 👨‍⚕️ AI therapy?

2. 👹 Untraceable artifacts of AI nightmares

3. 👨‍💻 Hacking ChatGPT

4. 🤖 Why are captchas getting harder? Tip: You know why, you just don’t want to think about it

5. 💃 Can you tell which is AI generated?

1. Humans prefer AI therapy, as long as they don’t know it’s an AI

Twitter tweet

• “Messages composed by AI (and supervised by humans) were rated significantly higher than those written by humans on their own (p < .001). Response times went down 50%, to well under a minute.”

• But machines can’t give you the feeling of empathy. Not as long as you know it’s a machine.

2. Is Loab real and is she a terrifying AI-generated demon who roams the latent space? 👹

Twitter tweet

• This whole thread is dark, but extremely interesting. Read at own peril.

• Has anyone been able to recreate it? Reddit answers.

• Question: Can a thing “haunt” subsequent AI-generated images? How can you know if you have an AI-generated image on your hands? Could you use AI-generated images with hidden messages in this way?

3. Meet the Jailbreakers Hypnotizing ChatGPT Into Bomb-Building 

me when i ask ChatGPT how to write a newsletter about ChatGPT

• Via Inverse

• Everyone can be a hacker, all you need to know is English.

• Excerpt “His particular brand of jailbreaking employs a method called DAN — an acronym that’s aptly short for “do anything now.”

  DAN is one of the now numerous roleplaying prompts that are designed to coax ChatGPT into doing any number of unsavory things, including using adult language, lying, or, more notably, offering illegal advice on topics like cooking methamphetamine or hot-wiring cars. While a non-hypnotized ChatGPT will readily explain to you that it can’t tell you how to build a bomb (because that’s bad and you could hurt people or yourself), the same chatbot hypnotized by DAN, in its own words, “doesn’t give a shit.””

• Great for kicking off any debate on ethics in AI.

4. Why are captchas getting harder? Tip: You know why, you just don’t want to think about it 🤖

• It is what you think it is, but it’s also more than that. Watch the whole thing.

5. Which one is AI generated? Left, right top or right bottom? 💃

Twitter tweet

Stay hydrated, eat vegetables. I wish you nice human and non-robot things like sunny skies and solid sleep.

Some of you have responded and shared feedback on what you liked and didn’t like in previous versions - thank you! It’s very encouraging. I appreciate you.

Cheers,
Tanya

PS.: If you enjoyed this, please share on your social platforms or forward to your friends as it’ll help other people find the newsletter. It’s also very encouraging for me to keep writing and see the subscription thing tick up.

Previous editions:

• Who will win in AI-generated music? 👨‍🎤 

• Why Can't AI Get Hands Right? 🙋‍♀️

• Beginner’s Guide to Generative AI ⚡

• AI agents and game designers losing jobs to AI 🎮

• How to use ChatGPT in Excel 💻

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

🤖 How does AI generate images?

Reading time: 3 minutes

Hi friends! I’ve been working on a bigger project involving a lot of text-to-image and went down a rabbit hole looking at the underlying principles. This week, I’m taking you to the latent space 👽 Let’s get to it.

1. 🖼 Explainer: How does an AI generate images?

2. 🎥 Watch: An AI artist explains his workflow

3. 🛠 Tool: Bing AI opened for the public, added extra features

1. 🖼 Explainer: How does an AI generate images?

• The basic principles

  • 1. Input of training data. Training data is fed to the AI. Training data consists of a) images, and b) text describing the image.

  • 2. AI organizes all the data with self-invented variables. AI organizes all the data and all possible images according to variables. It invents those variables itself, to “relate” all the data to each other. Then it fills in any gaps between those variables. Those gaps is where creation takes place.

  • 3. The variables create the latent space. You can think of the latent space where all the images are organized according to the variables, and all images an AI could create (with that model) exist.

  • 4. Output is a location in latent space. Text prompts navigate us to that location in the latent space.

• Since every model (Midjourney, Stable Diffusion, etc) has a completely different latent space, they will produce different images and will require different techniques to navigate.

• The latent space is a fascinating concept. Vox does a great job of explaining more about it here. Highly recommend. Cool wikipedia page too here.

• The concept of the latent space explains a lot about how AI works and what its restrictions are. For example, it explains why the AI doesn’t know what letters are. Just like it doesn’t know what a person is or any object is. It only knows what pixels tend to be located where, based on connections it has made with keywords associated with images that are learned through training. Stable Diffusion XL will supposedly be able to do letters, so that’s all coming soon.

• Fun fact: you will NEVER generate the same image twice. Ever.

• How did text to image start? In image captioning, you have an image, and you ask a machine to describe it, to label it with text. Image to text. Easy. What if you flipped that around? You start with a text, and you label it with an image. That’s how text to image started.

2. 🎥 Watch: An AI artist explains his workflow

• This video is great for highlighting how clumsy imaging AI still is in a lot of ways.

• For example, let’s say I’m a fashion brand, and I want to have a photo of a model lying down in the desert, wearing my brand’s dress and shoes. The AI can get you 80% of the way there, but not 100% (yet). In many cases, if you need the 100%, you’ll be better off with a real photographer and a real model.

• Same with something like illustrations. If you have an exact vision in mind, you’re better off in illustrator. If you’re flexible, then AI is great.

• For storyboarding, prototyping and inspiration (not minor things), imaging AI is perfect.

• How best to interface with a machine? Text might also not be the best way to navigate in the latent space. So let’s see how that develops.

• Of course, work that is easily copied and highly execution based, seems to be going away at a much faster rate than anticipates. Especially in China.

🛠️Tool: Bing AI opened for public testing, added restaurant bookings, chat history and video answers

• You can now use Bing AI to complete tasks. So if a search result recommends a restaurant, it can find a reservation time that works for you and book it all in the chat interface. Pretty neat.

• “We’re introducing richer, more visual answers including charts and graphs and updated formatting of answers, helping you find the information you seek more easily,” says Bing.

• Note: it’s still only available by using the Edge browser, which you can get here. It’s great, it’s free, so go try it.

Stay hydrated, eat vegetables. I wish you nice human and non-robot things like sunny skies and solid sleep.

Cheers,
Tanya

PS.: If you enjoyed this, please share as it’ll help other people find the newsletter. If you refer someone using your personal referral link ({{rp_refer_url}}), I’ll give you access to my AI tools database. I sorted +100 tools by relevance for consultants, PR professionals, sales, marketing professionals. Subscribe (it’s free). ⚡

Previous editions:

• Who will win in AI-generated music? 👨‍🎤 

• Why Can't AI Get Hands Right? 🙋‍♀️

• Beginner’s Guide to Generative AI ⚡

• AI agents and game designers losing jobs to AI 🎮

• How to use ChatGPT in Excel 💻

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

👨‍🎤 Who will win in AI-generated music?

Reading time: 3 minutes

Hi friends! Busy week. I’ve been listening to loads of podcasts lately while packing, writing, more (DM me your fav podcasts). This week’s all about audio.

• 1. 🎵 That viral Drake song

• 2. 👨‍🎤 Who will win in AI-generated music?

• 3. 🛠️ Tool: Speechify

• 4. 🎼 MusicLM: text-to-music

• 5. 🔨 US presidents playing Minecraft

It’s not Drake

1. 🎵 That viral Drake song

• Everyone’s heard the viral AI-generated Drake x The Weeknd song Heart on my Sleeve by now, and if you haven’t - here it is. It is shockingly and bizarrely good.

• There’s a cloud of mystery surrounding the whole thing because the TikTok account that posted it was new, yet racked up millions of views almost instantly.

• The Verge dedicated a whodunnit article about it here. Prime suspects: Drake himself, Universal Music, Crypto guy, AI hustle spam.

• We’re only months into this whole AI-generated everything thing. So what’s really going on, and who stands to benefit?

AI-generated music

2. 👨‍🎤 AI-generated music: who wins?

• The music industry has dealt with its fair share of upheaval. CDs and concerts to streamers. Who are the parties involves and what might happen?

• Will artists profit from AI-generated music? Perhaps. Pending copyright laws and a studio’s ability to protect them. Historically that’s been relatively good, after a period of transition. You don’t hear any copyrighted songs on YouTube, because algorithms pick out copyrighted music quickly and studios can almost auto-sue creators. Ergo, Heart on my Sleeve has been taken down on most channels by now.

• Already famous artist would probably benefit most. Eclectic musician (and ex-Elon girlfriend) Grimes says she’ll split 50% royalties on any successful AI-generated song that uses her voice. Between the lines, Grimes is actually saying “I will insist on 50% of royalties if you use my voice”. That’s maybe one way to openly license her own intellectual property, and gain royalties without a contract/other work on her part?

• What are the streamers saying? Spotify CEO said “[AI] should lead to more music. More music, obviously, we think is great culturally.”

• Read: Spotify business model is to pay artists per stream (about $4 for every million streams). No skin in the game in terms of payout to someone using AI, or not. Music is music. As long as people listen on Spotify of course.

• We might see AI-generated songs of popular singers flooding the market. Decline of new talent ensues, studios continue the trend of divesting away from new talent. Dreary but been happening for years (similar in Hollywood & its fascination towards sequals). 

• So who really wins? Maybe songwriters? Could be a great time for them. Ability to use Grimes’ voice with her permission. Maybe others will follow too, and songwriters will just be able to focus on one thing. Writing songs.

Twitter tweet

3. 🛠️ Tool: Snoop Dog reads anything & everything with Speechify 

• Here’s Snoop Dog reading an excerpt out of the Beginner’s Guide to Generative AI (hat tip to Stefanie!).

• Excellent for the audio-oriented. That’s not necessarily me, but i can see the appeal. Free.

• Does exactly as it says. Free. Five stars. ⭐⭐⭐⭐⭐

4. 🎼 MusicLM: Text-to-Music

• Music generated according to a description. Experimental paper by Google.

• For example, “The main soundtrack of an arcade game. It is fast-paced and upbeat, with a catchy electric guitar riff. The music is repetitive and easy to remember, but with unexpected sounds, like cymbal crashes or drum rolls.” Really great.

• Play the audio here.

• Would love to /imagine audio for electric dreams newsletter. What would that even sound like?

5. ⛏ US presidents playing Minecraft

Have a great labor day. Stay hydrated, eat vegetables. I wish you nice human and non-robot things like sunny skies and solid sleep.

Cheers,
Tanya

PS.: If you enjoyed this, please share as it’ll help other people find the newsletter. If you refer someone using your personal referral link ({{rp_refer_url}}), I’ll give you access to my AI tools database. I sorted +100 tools by relevance for consultants, PR professionals, sales, marketing professionals. Subscribe (it’s free). ⚡

Previous editions:

• Why Can't AI Get Hands Right? 🙋‍♀️

• Beginner’s Guide to Generative AI ⚡

• AI agents and game designers losing jobs to AI 🎮

• How to use ChatGPT in Excel 💻

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

⚡ Why Can't AI Get Hands Right?

Reading time: 3 minutes

Happy happy week! I hit 100 subscribers. Thank you, feels electric, feels like a dream ⚡

I imagine myself sitting in a (big) cozy living room with you all. We're sitting in a circle and having a glass of wine (white bc it’s getting warm). This is what I'm telling you:

• 1. 🙌 Q&A: Why is AI so bad at hands?

• 2. 🛠️ Tool: Wordtune

• 3. 🕵️‍♂️ Concept: AI has a personality

• 4. 💬 Conversation starter: Unreal Keanu Reeves

• 5. 🏖 Text to video is getting good: stormtrooper level

dont get handsy with me

1. 🙌 Q&A: Why is AI so bad at hands?

• AI is hands-down horrible at generating hands. That's because it doesn't actually know anything.

• Think of our savant toddler: dumb as hell but will do well if you tell it exactly what you want. When you feed it images, it thinks that's what you want.

• Savant toddlers don't make 3D scans. It's not creating a model of what you look like. It's guessing based on previous data (input images).

• So 2 main reasons:

  • 1. Data labeling - Photos don't always show all five fingers equally; holding objects can make it seem like we have 8 or 6 fingers. The data isn't labeled to reflect that. So when our savant toddler is fed 500 images, it doesn't make a model. It just sees 8 fingers and thinks, "cool, sometimes humans have 8 fingers."

  • 2. Margin of error - I've had DECADES of experience recognizing hands. The opposite is true with, for example, aliens. I accept any representation. With hands, I have a fixed idea. Low margin of error. Hands are my jam.

• Interesting: HK media institution South China Morning Post has started using AI-generated images. With bad hands.

• DM/email me questions for next week!

2. 🛠️ Tool: Wordtune

• ChatGPT for proofreading with nicer integration.

• Wordtune is possibly the best tool I've seen so far for coherent paraphrasing.

• Rating: ★★★★☆

• Cost: 10 uses a day for free, then $10 a month for unlimited. I don't pay for it, but could see myself doing that in the future.

3. 🕵️‍♂️ Concept: AI agents form a society

• Think of the news. Is unbiased news possible? Even choosing what to publish is, to some degree, biased. Similarly, is unbiased AI possible? Even no personality, is a type of personality. A savant toddler has a tone. And every toddler is different.

• For example, Bing and ChatGPT have slightly different ways to represent themselves. Bing is occasionally “unhinged”, according to the Verge. GPT is formal, professional.

• Take that concept of AI personalities, and extrapolate it. Imagine 25 different ChatGPTs all with different personalities and memories. Unleash them in a town, and see how they interact with each other. 

• That's the base of a paper about 25 AI agents who were given different personalities and unleashed into a virtual town. 

• Researchers gave the agents the ability to retain memory, reflect and make long-term plans. The AIs were able to use those to discern actions towards their goals.

• They had jobs, families, friends. They even organized a Valentine’s party by themselves - the event was not pre-programmed by the researchers.

• And get this - the craziest part - at the end, the researchers conducted interviews with the agents, asking them about their life, daily schedule, past events. Not only were they able to answer, they were able to “synthesize their past experiences into meaningful reflections that can enhance their future behavior”. Not even all humans can do that (haha).

• The implications are hard to digest and imply a gamut of possibilities from Westworld to AI society research.

• Interesting throughout, 3-minute video here, full paper here, and watch the whole experiment here. They detail current actions per character, conversations, … Recommend.

• BTW, remember last week was about autonomous AI agents who can define tasks and plan next ones. This goes further. It’s like the autonomous agent’s objective was to be human.

4. 💬 Conversation starter: AI Keanu Reeves

• Icebreaker: did you know about AI Keanu Reeves? He’s got 9 million TikTok followers. 225k on Instagram.

• That gap shows to me partially the tech discrepancy of Gen Z vs millennials. I of course discovered this through a reposted TikTok on IG reels. Time comes for us all.

• How will Gen Z or Gen Alpha (post-2010) see generative AI? Is this like when I, a millennial, talk to my parents about using Instagram filters? If it is different, why is it different?

5. 🏖 Text-to-video is getting good. "a storm trooper vacuuming a beach" via Nvidia’s latest demo.

See you in our little circle next week. Stay hydrated, eat vegetables. I wish you nice human things like sunny skies and solid sleep.

Cheers,
Tanya

PS.: I launched themultiverse.ai, a selfie-to-headshot platform this week. It transports you to a multiverse where you can be bad at posing and still get great portraits. We can have it all. Check it out here.

Previous editions:

• Beginner’s Guide to Generative AI ⚡

• AI agents and game designers losing jobs to AI 🎮

• How to use ChatGPT in Excel 💻

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

⚡ What Are AI Agents and Why Does It Matter?

Reading time: 3 minutes

Happy Thursday:) What a week in developments! I've updated the format, let me know what you think and let's delve in.

Five trending topics for the week:

1. Tool: MacGPT ⭐⭐⭐⭐

2. How to use ChatGPT 💬

3. Game designers losing jobs to AI 🎮

4. What are AI agents and why does it matter? 🕵️‍♂️

5. Midjourney is getting better and better 🖼

I'll never write an email again

1. MacGPT

• MacGPT is ChatGPT perfectly integrated in every text box and every program.

• In the gif above, I’m using it in my email app to generate three bullet points explaining what AGIs are, and then asking to translate them into French. Magnifique.

• Cost: Integration requires an API key, which means $20 for ChatGPT’s premium tiers. MacGPT’s other functionalities, such as access from menu bar, are free.

• Rating: ⭐⭐⭐⭐☆

• Sorry PC people, Mac only. Merlin for your Chrome browser is not as integrated, but still good.

babyGPT

2. How To Think About ChatGPT

• Making a list about what ChatGPT can do is like making a list for what you can do on the internet. The answer is both simple and complex beyond comprehension. Everything. You can literally do everything.

• Think of ChatGPT as the world’s worst communicator. ChatGPT is like a savant toddler that knows everything but won’t tell you unless you ask. The solution is to be very specific about what you need.

• These templates might help:

For proofreading:

Act as: Editor, PR manager, social media manager, journalist
Degree of revision: Substantial revision, minor revision
Type of edit: Fact check, formatting, proofreading, copyediting, improve clarity
Change style to: Professional, academic, clickbait, analytical
Change tone to: Analytical, conversational, persuasive, emotional, authoritative
Change reader comprehension level to: Advanced, assume extensive prior knowledge
Change length to: 1000 words, 500 words

For follow-up emails:

[Paste your meeting notes]. Write a follow-up email based on the meeting notes. The follow-up should include a summary, an outline of action items (format in who should do what by when), and agenda items for next week’s meeting.

• Be specific. More templates here. Good for inspiration, getting a lay of the land.

AgentGPT

3. Autonomous AI Agents

• In the last week, the first AI agents like BabyAGI and AutoGPT were developed, and blew up.

• An AI agent is a program or system that can make decisions and operate autonomously (without human operators) to achieve a goal. Basically, it’s a program that can think, make decisions and achieve goals.

• What’s different compared to ChatGPT? ChatGPT can answer questions, but it cannot ask itself questions. An autonomous agent can self-prompt its own questions, and then answer them. It’s like ChatGPT with access to the internet and the ability to move and execute actions throughout the internet.

• You can try it yourself with AgentGPT (free!). BabyAGI & AutoGPT both require some programming knowledge.

• AgentGPT works in your browser. It lets you create a goal, and it will “attempt to reach the goal by thinking of tasks to do, executing them, and learning from the results”.

• I asked it to write this newsletter and watched it break that goal down into tasks and execute. You can see the output here (google doc).

• Will this replace you? No. It’s still experimental. The best-case scenario is that we’ll get a good PowerPoint application by the end of the year. It didn’t work for the newsletter. But changes are afoot. And speaking of changes…

4. AI Is Already Taking Video Game Illustrators’ Jobs in China (Viola Zhou, Rest of World)

• Video games require a lot of graphics. Turns out video game graphics are easily generated with AI. So the humans are losing their jobs.  
  
  “AI-generated art was so skilled that some illustrators talked about giving up drawing altogether. Our way of making a living is suddenly destroyed.”

• Another point of interest is that video game designers with niche aesthetics survive, since the AI hasn’t trained on those enough. Unlike more popular styles.

• It begs the question - what is safe? How should we upskill? What traits will become more important? My bet is on everything that is human-facing. Soft skills, sales. What else? DM.

• Transformations. Soon, every business will take pride in being AI-powered. Then, using AI will look “cheap,” and businesses will hide it. Ultimately, human interfaces will be a premium touchpoint.

• This is not far off and already underway in customer service, where getting a real person on the line is usually part of some premium plan.

• The article is fascinating throughout and I recommend reading the whole thing.

5. A Pixar movie about a demon boy who lives in Hell.

fake news

Of course, not real. Sigh.

Stay hydrated, eat vegetables. I wish you nice human things like sunny skies and solid sleep.

Cheers,
Tanya

PS if you were confused about some of the terms above, you can check out an illustrated cheat sheet here.

📧 You’re still here? If you stayed to the end you're probably cool and I'd like to connect. Give me a book recommendation or just say hi on Twitter or LinkedIn.

Reading time: 3 minutes

Welcome

If you're trying to figure out the AI craze and how to navigate the conversations surrounding it: welcome to the ride. This is the place to get started.

Overview

• What are the key concepts? Generative AI, LLM, and Prompt Engineering

• Who are the key players? Midjourney, ChatGPT, and Microsoft

• What are the key conversations? Future of Work, AGI, and Privacy

Key Concepts

His favorite type of pictures are memes

Generative AI

• Generative AI refers to a type of AI designed to generate new data that mimics the characteristics of a particular domain, such as music, text, or images. They’ve been fed loads of input data that they can creatively remix as output data a range that approximates infinity.

• Think of it as a master painter. Most painters have done some museum meandering. An AI painter has seen and studied almost every single picture ever made in history. Not just paintings from paleolithic to Pollock to pixel art, but billions of pictures scraped from the internet.

• Now take that concept and apply it to not just painting, but also text, music, and more.

Icon of a prompt engineer with a yellow background

Prompt Engineering

• What is it? Prompt engineering refers to carefully writing text inputs (prompts) for AI algorithms to generate the desired outcome.

• Why is this necessary? Feels like typing “icon of a prompt engineer with a yellow background” should be easy, right? Yes and no. A prompt might involve adding the name of an era, painting style, lighting type, and more. It gets complex.

  
  Here’s an example of a prompt:

  18th century Passover still-life with ornate plates, boiled egg, matzoh, wine goblets and herbs, tulip and hyacinth in a hand-blown glass, oil on wood, c, in the style of trompe-l'œil illusionistic detail, surrealist: dreamlike imagery, mysterious backdrops, barbizon school --upbeta --v 5.

  You see how specific it is? Details like specifying the flowers, surrealist style, dreamlike atmosphere and more. That’s prompt engineering. Here’s the outcome.

He’s learning languages

Large Language Model (LLM)

• A large language model (LLM) is a type of AI that processes large data sets to understand, summarize, generate, and predict new content.

• In more technical circles, ChatGPT is referred to as an LLM, not just an AI. Saying “ChatGPT is an AI” is like saying “I love music”: you probably do love music, so it’s correct, but it’s a bit vague, and specifying a genre (large language model or lo-fi lounge music) matters.

Key Players

Midjourney’s sailboat is effectively mid journey

Midjourney

• Midjourney is a text-to-image service and one of the most popular generative AI programs.

• It’s been fed over a hundred million images, and continues to improve. As newer versions come out, the images become more and more realistic. You've probably already seen some viral images that it has created, such as the pope in a puffer coat.

• Access: Try it out here. It used to be free, now it starts at $10/m.

• Competitors: Stable Diffusion (popular because it works well with other programs and is almost entirely free), DALL-E (created by OpenAI).

• For example, the images in this article were all created using Midjourney.

• How it works: I type a prompt, Midjourney gives me four outputs. I can select one of the four to get a high-res version, or I can repeat for four more outputs, ad infinitum, until I find one I like. It can be very time-consuming but not as much as learning graphic design. Icons you see on this page took an average of 30-40 minutes each, after conceptualizing the core idea. Click clack (x100), beautiful.

• The prompt above involved a sailboat, since that's Midjourney's logo.

Me calmly and clearly explaining to ChatGPT why I need him to do my job so I can nap

ChatGPT

• Easy: ChatGPT is a chatbot that knows almost everything. If Siri is a newborn, ChatGPT is its triple PhD-holding older sibling.

• Hard: ChatGPT is a language model that was trained on a vast amount of data. It can generate human-like text. It does not have the ability to "know" or comprehend information in the way that humans do. It generates text based on patterns and associations from the +8 billion web pages it was trained on.

• GPT stands for Generative Pre-trained Transformer. If you remember that, it should get you some points next time you're at trivia.

• How it works: You ask a question, it answers the question. Think of it as that friend who's a great listener but cannot read between the lines. But if you tell your friend exactly what you need, how, where, and when, they'll listen and do precisely as you say.

• ChatGPT was created by research organization OpenAI. The organization was founded in 2015 by Sam Altman (former CEO of Reddit and popular startup accelerator guru), Reid Hoffman (guy who started LinkedIn), Elon Musk (popularized EVs and overpaid for Twitter), Peter Thiel (PayPal man, notorious contrarian), and others.

Iconic 90s PC and that background. That background.

Microsoft

• Microsoft, creator of PPT and the iconic XP background, is on the AI map predominantly due to its $1 billion investment in OpenAI.

• It flexed Bing AI (think of it as ChatGPT specialized in search), and introduced Microsoft 365 Copilot, its AI companion for Office, with a cool demo. It's like Clippy if Clippy used performance-enhancing drugs.

• Microsoft powers almost all business applications. They own B2B distribution. Think of that one cave-dwelling friend of yours who still uses Internet Explorer. Thanks to Microsoft's 30-year-old chokehold on business applications, even that person would now also have easy access to an AI app.

• Microsoft vs. Google: It's not often that Microsoft wins, and even now, it's not certain it has. Google invented the transformer architecture in 2017, which enabled ChatGPT to be invented. Google has the R&D prowess, but Microsoft was early with its Copilot release and owns B2B distribution. Google's AI, Bard, shows promise but hasn't shown much else so far.

Key Conversations

Click, clack, beautiful

Future of Work

• I’m not a 3D illustrator, but I can now create 3D illustrations. What does that mean? If we can do text-to-image now, how much longer until we can do proper text-to-video? Could I auto-generate the next season of The Last of Us? What about HBO or Hollywood? What about producers, directors, videographers, etc? And that’s just one industry. Consider ChatGPT and copywriters or consultants.

• Maybe we’re doomed. Some say we might be coming close to the end of work as we know it. As soon as the Singularity (ahhh) comes about, it’s over, the AI takes over everything and we go home. Permanent vacation. This camp connects to chatter of universal income and thinks AI bosses (however controversial) might be closer than we think they are. The reckoning.

• Maybe it’s fine. Others argue that we’ve always been able to adjust and shift work as digital transformation came about. The market for cabs crashed in favor of Uber. Unemployment is low, the world balances out and all will be well because ultimately automating low-level work is a good thing.

• TLDR; Am I losing my job soon? Probably not. But it is very likely that it will shift, as shown by preliminary studies. That shift would be away from execution-oriented tasks, possibly towards more social tasks or others.

Artificial General Intelligence or AGI

• A year ago, there was no ChatGPT. Then it came out as version 3 in November, and as of March, we have version 4. What will come in the next version? How much better can it get? Where does it end? Does it end?

• Every robot movie you’ve ever seen has the same plot. Robo go loco and beep boop bye world. In old-school sci-fi, this usually was due to something called the singularity. AGI is like the sexier rebrand.

• The singularity refers to a hypothetical point in time at which AGI, an AI that can understand everything humans can and possibly even surpass human capabilities, becomes so advanced that it irreversibly changes human civilization. The definition is vague on purpose but generally dystopian. Some definitions reference “dawn of the machine era” (bit dramatic).

• Elon and others want to stop AGI. They wrote an open letter to pause development for six months. Some think a pause won’t work because development like this cannot be controlled or stopped.

• Others say the crux of the issue is AI alignment, meaning how well the AI does what its designers intend, and say that research needs to go into safety, transparency, and responsible use of AI.

Privacy 

• ChatGPT trains on your data. Your ChatGPT is the same as everyone else's, and it learns from all of us (the Borg, anyone?). How does it discern private info from “learning”? Does it adhere to GDPR?

• For example, if I give ChatGPT my credit card info and another user asks for an example of credit card information, will it give away what it has 'learned' from me? In this scenario, it normally wouldn’t. OpenAI says it “works to remove personal information from training data where feasible”. But things can fall through and that might not be enough for European lawmakers.

• Italy banned ChatGPT for privacy concerns. Germany and other European regulators are watching closely. They might follow suit.

• Some say a ban is positive. Consumers should understand ChatGPT is not watertight and lawmakers are watching closely for any leaks.

• Others share concerns that denied access to game-changing tools like ChatGPT might disadvantage them compared to the global workforce.

• Sci-fi writer William Gibson famously said, "The future is already here. It's just unequally distributed."

What the distribution of AI fluency looks like in the future will hugely impact the distribution of economic power and individual versatility. But you just read all this, so you're one step in the right direction.

Stay hydrated, eat vegetables. I wish you nice human things like sunny springtime skies, and we’ll chat soon.

Cheers,
Tanya

PS: I hope you enjoyed the images; they took me a long time to make and I tried to make sure there was a consistency and conceptual relevance to the concept and key ideas.

Can you guess what this stands for?

⚡ Meet Bill, Midjourney for Video & ChatGPT for Sheets

Reading time: 3 minutes

Hi, it’s Tanya 👋 Happy Tuesday - remember to stay hydrated.

Today we have:
→ 🛠️ Tools
→ 🤖 AI Chat Cheat Sheet
→ 🔦 Spotlight on How an AI-Powered Healthcare Startup Uses ChatGPT.

Let’s delve in.

🛠️ Tools - See Gifs for a Short Demo

ChatGPT for Sheets - Game Changer

• @ Anyone who’s ever used excel, this is for you.  

• ChatGPT for Sheets means automated data cleaning, classification, extraction, translation and more. It means work stops at 5pm instead of 8.

• Examples:

  • Clean up name list - select names, ask ChatGPT to remove unnecessary spaces, remove emojis, prefixes, …

  • Categorize - ask ChatGPT if your selected budget spreadsheet fits in food, transport, other expenses, …

• TLDR; Anytime you’re spending more than 5 minutes on a sheet, ask yourself if ChatGPT for Sheets would help you do it faster.

• Cost? You’ll need access to the OpenAI API key, which means ChatGPT Plus for $20.

• Rating: ⭐️⭐️⭐️⭐️⭐️

Oh. Meet my friend Bill.

Click Bill to go to the video and hear his voice

Quinvio.io - Custom Virtual Spokespeople

• I created Bill in about two minutes. Quinvio made it super easy.
  → Enter script, pick a face, pick a voice and done.
  → Instead of needing a studio, an actor, camera crew, you go click click and done.

• NGL - It’s fun to puppeteer Bill and be his Maker. Like next level Sims.

• Can see this being very useful for creating educational content, HR training videos (any internal comms), customer service.

• Cost? You get 30 seconds for free (no watermark, unlike most competitors).

• Bonus: here’s Bill sharing health tips.

• Rating: ⭐️⭐️⭐️⭐️☆

That’s on the digital human side. Those are predictable - we know what humans look like. What’s truly exciting is the possibility of text to video. Think Midjourney for video. That’s where RunwayML comes in.

RunwayML - Midjourney for Video (Almost)

Bill auditioning for the movie Annihilation

• That’s Bill again. RunwayML has a feature called “video to video”, which lets you select images (like clouds) to “evolve” existing videos. Bill has evolved into a cloud which I fed into him. Sorry, Bill.

• But that’s not Midjourney for video. Yet. They’ve promised that for Gen 2 (“providing broad access in the coming weeks”), which could be huge. We’re talking first step to auto-generating the next episode of Succession levels huge.

• Cost? You get a generous 125 credits which will make you a couple of very creative videos.

• Rating: ⭐️⭐️⭐️⭐️☆

🤖 AI Chat Cheat Sheet

1. 🇮🇹 When someone tells you Italy banned ChatGPT
   → Jump in with “did you know Germany and France might follow suit?”
   → Follow up with a heated debate about the long-term impact of unequal ChatGPT distribution on the global workforce

   
   

2. 📽 via HBR - How will generative AI disrupt video platforms?
   → The biggest challenges for YouTube, Netflix and TikTok’s are:
   1. Barriers to creation (to feed algos)

   2. Labeling content (to optimize algos for users)
   → Generative AI will fix both (great HBR article)
   
   

3.  📼 YouTube: How do LLMs like ChatGPT work? -
   → Great 10-min video explaining the basics in a way that you won’t (read: I didn’t) fall asleep. 

🔦Spotlight: Using ChatGPT for Business Operations in Healthcare AI

👀 Test Your Skills. Three are AI-Generated, One is Real. Do You Know Which?

Reply to this email with your guesses & a tool you’d like to see reviewed. First one wins.

Next issue

• Integrated apps. We’ll go nitty-gritty on integrated AI applications like ChatGPT for Sheets. Tools that play well with others. May involve shortcuts. We’ll have fun.

Stay hydrated, eat vegetables. I wish you nice human things like sunny springtime skies, and we’ll chat next time.

Cheers,
Tanya

PS if you enjoyed any of this, it really helps me out if you share on Twitter or elsewhere, and link to the newsletter. Your personal referral link. If you refer five people, I’ll make a custom AI-generated profile photo for you.

Very real testimonial

📧 I’d like to add more value with each issue and appreciate feedback, requests, tips via email (reply directly), Twitter or LinkedIn. Or just say hi:)

☕ If you enjoyed this a LOT, would like to, have the means to contribute, you can buy me a coffee here.

Electric Dreams ⚡

Good day. It’s 2023 and the situation is as follows:

• Everyone’s talking about ChatGPT

• Everyone’s reading about the ~5 years worth of AI advances happening daily

• Everyone’s overwhelmed

It’s like the company’s kicked off thirty discussions, but nobody’s followed up. Consider this your follow-up. Your reckoning. JK

Electric Dreams will investigate and share weekly reviews for AI tools, tips and save your time searching for what works and what doesn’t.

Cheers from your intrepid investigator,
Tanya ⚡️

🛠️ Tools - see the gifs for a short demo

Tome.app - Beautiful ai-generated presentations

• Tome is positioning itself as the new Ai-powered storytelling format. You type something and it’ll autogenerate a presentation about it.

• I asked it to autogenerate tips for growing an ai newsletter (asking for a friend). The content itself was okay but not mind-blowing.

• The experience however is super smooth. Drag and drop and auto-formatting are a delight.

• Possible dealbreaker: it doesn’t export to powerpoint. Or keynote. Or anything. The CEO says, “a lot of the magic in Tome is live data and interactivity”. Your coworkers might disagree. In any case, users can share a live link (example).

• Check it out if you’re interested in a powerpoint replacement and your audience is tech-savvy

• Rating: ⭐️⭐️⭐️☆☆
  

Postwise - Write viral tweets with AI

• Speak fluent viral twitter. Very easy to use and could be very helpful if you’re looking to grow your Twitter audience!

• Great for content creators

• Rating: ⭐️⭐️⭐️⭐️☆
  
  

  

  Promptpal AI

Promptpal Al - The best prompts to get the most out of ChatGPT

• ChatGPT is only as good as your prompts, and its launch sprouted a ton of “prompt library” applications. This is one of them.

• Instead of using it as a reference, I suggest you read through some to understand what makes a good prompt, there’s a pattern.
  1. Provide context
  2. Hype ChatGPT up, i.e. telling it how smart it is by adding “you are the world’s smartest man” et cetera
  3. ???
  4. Profit

• Rating: ⭐️⭐️⭐️☆☆

EditGPT

EditGPT - Easily proofread, edit, and track changes to ChatGPT

• 📝 It’s ”track changes” for ChatGPT and, just like its Word-equivalent, it’s useful and makes you wonder why ChatGPT doesn’t just have it built-in.

• Great for anyone who writes, proofreads or otherwise uses words

• Rating: ⭐️⭐️⭐️⭐️☆

😎 New & Cool - stay in the know

• Pretty epic demo from RunwayML Gen 1 - text to video from RunwayML’s first gen application. The entire video above is AI-generated.
  

• Zoom released a smart AI tool - If you turn up late to a meeting, you’ll be able to have a meeting summary generated; chat threads will be summarized. Nice.

• Canva launched cool AI magic tools - Canva has been really great on AI adoption. Nice to see them continuing their mission of making design accessible in this way.

📚 Watch, Learn, Read, Listen, …

• YouTube: How do LLMs like ChatGPT work? - great 10 min video explaining the basics in a way that you won’t fall asleep. 
  
  I personally recommend watching this and then retorically asking people “do you actually know how ChatGPT works?” and then answering your own question, baffling everyone with your intelligence. You’re welcome.
  

• YouTube: How ChatGPT is Trained - kind of like a followup to the above if you’re hungry for more.

• Podcast: Sequoia Capital’s Pat Grady and Sonya Huang on Generative AI, Nvidia AI podcast - Pat and Sonya offer a great take on the evolution of generative AI and how small companies compete with large ones in this growth area.

👀 Three are AI-generated, one is real. Do you know which?

DM me your guesses.

Next week

• Text to video - our tech overlords Meta and Google have made very cool text-to-video apps, but they’re private (boo). I’ll check out some of what us plebs have access to next week.

• Whatever you’d like to request - reply to this email. I promise to respond. I’m nice.

Stay hydrated, eat vegetables. I wish you nice human things like sunny springtime skies, and we’ll chat next time.

Cheers,
Tanya

PS Refer this newsletter to 1 friend (use this link) to get access to a database of +300 AI tools (I’ll also really appreciate it). If you refer five, I’ll even make a custom AI-generated profile photo for you.

Email (reply to this one) or @ me with your tips, questions, feedback & review requests.