Building and Creativity �in the Age of AI
The Lead Conference 2024
New York, NY
Disclaimers
�Thank you.�- Emmett
Introduction
I’m Emmett Shine.�
I helped found:
Introduction
I’ve worked at the intersection of design and tech for 20 years.
�Always tinkering with software and tools.
�I try to be a translator / conduit �between subcultures, niches, emergent technologies and mainstream adoption.
Introduction
So, this is a perfect topic to try my hand at.
My hand
Introduction
I’m a Libra. ⚖️��So, I’m going to try and build �a balanced conversation on AI.
Was drinking this while writing; felt fitting
Introduction
Like most ‘creative directors’ or �‘non technical founders,’ I am �by no means a real expert here.
It’s important to clarify these are �my opinions and to make your own judgement calls.
Stock or AI?
Introduction
There are three parts to this presentation:
Part I: How to Think About AI� A. How to Think About the Future� B. How to Think About Technological Change� C. The Creative Process
Part II: 1 Year Ahead� - How to use AI today
Part III: 10 Years Ahead� - How we’ll use AI down the road
Stock or AI?
Part IA: How to Think About the Future
Part IA: How to Think About the Future
Part IA, is going to deal with how humans can have irrational responses to the unknown, aka the future.
Some of us over index on hope, or fear, when presented with something we don’t understand, aka the future.
My goal is to attempt to normalize how we think about the future; helping us not have too much hope, nor too much fear.
Part IA: How to Think About the Future
Let’s first start with why we sometimes place too much value, potential, or hope, in that which is new.
Part IA: How to Think About the Future
As I was preparing for this, I finished the book Atomic Habits. The last sentence of the last page had this quote:�
“New plans offer hope because �we don’t have any experience �to ground our expectations…
Youth is easily deceived because �it is quick to hope. There is no experience to root the expectations in. �In the beginning, hope is all you have.”
Part IA: How to Think About the Future
TL;DR:
We hope when we don’t know enough.
We only become more tempered through experience…
We’ll get into how to gain experience, so as to not fall victim to hopium, later.
Part IA: How to Think About the Future
The second part of this section, �is about how we also often lean into fear when confronted with change, uncertainty, or the unknown.
Part IA: How to Think About the Future
Here’s a simple, but appropriate, quote:
"Fear always springs from ignorance."
Ralph
Part IA: How to Think About the Future
I wanted to share this excerpt from The Mountain is You. �Most of the audience reading this are not as prone �to a fear of technology, but instead may think they �can outsmart the technology to come:�
“Instead of trying to use your intelligence to hack what’s next, �try to get better at where you are currently. That’s what’s really going to change the outcome of your life.”
Part IA: How to Think About the Future
TL;DR:
Part IA: How to Think About the Future
Hope vs. Fear��My proprietary, not totally accurate, �Venn diagram for where you want �to be for thinking about the future. �Pull from your past experience.���“History Doesn't Repeat Itself, but It Often Rhymes” ��– Mark Twain
Part IB: How to Think About Technological Change
Part IB: How to Think About Technological Change
"You can't connect the dots looking forward; you can only connect them looking backward. So you have to trust that the dots will somehow connect in your future."
— Steve Jobs, Stanford Commencement Speech, 2005
Part IB: How to Think About Technological Change
To prepare for the future, it’s important to sometimes look backwards. ��There is more to learn from �the past then we often give credit.
For this section, I hope to establish �an understanding that AI is part of a long lineage of technological change.
One that should help us plot where things might go.
Yayoi Kusama
Part IB: How to Think About Technological Change
I wanted to share some frameworks that directly relate to technological change.
This section was inspired by Farnam Street’s excellent essay, �‘Gates’ Law: How Progress Compounds and Why It Matters’
Part IB: How to Think About Technological Change
Gates’ Law:��“Most people overestimate what they can achieve in a year, �and underestimate what they can achieve in ten years.”��- Roy Amara, Stanford computer scientist, 1960s
Part IB: How to Think About Technological Change
Similar to hope and fear, humans overestimate what happens in the near term, and underestimate what can happen down the road.
Here’s a helpful graph:
Part IB: How to Think About Technological Change
Here I put an overlay of �a green line to represent where a given technology actually is (versus where the expectations are).
0
5
10
15 years
Part IB: How to Think About Technological Change
Gate’s Law will be our framing for ‘Branding & Creativity in the Age of AI’:
Part IB: How to Think About Technological Change
Some more helpful technological laws…
Part IB: How to Think About Technological Change
Moore’s Law:��“The number of transistors on an inch of silicon doubles every eighteen months.”��Gordon Moore, co-founder of Fairchild Semiconductor and Intel
Part IB: How to Think About Technological Change
By examining Moore's Law, which predicts the exponential growth of computational power, we can frame the seemingly rapid advancements in AI �as a natural continuation �of long-term technological progress.
�
Part IB: How to Think About Technological Change
The Adjacent Possible:��Each new innovation adds to the number of achievable possible future innovations. It opens up adjacent possibilities which didn’t exist before, �because better tools can be used to make even better tools.��- Stuart Kauffman, medical doctor / theorist
Part IB: How to Think About Technological Change
‘The adjacent possible' highlights how each technological advancement opens the �door to new innovations.
We can better grasp how AI �will lead to new progressions, �which can help us view its future trajectory as a natural and pragmatic extension of historical technological progress.
Stock or AI?
Part IB: How to Think About Technological Change
The illustrations on the following pages by Simon Hørup Eskildsen do a fantastic job of showing how seemingly disparate cultural and technological developments lead to unforeseen innovation.
Looking at AI’s path forward, we can see our next year a bit clear, but it’s harder to imagine 1:1 what is ahead in the ‘shadow future.’
‘the shadow future’
Part IB: How to Think About Technological Change
Illustrations courtesy of Simon Hørup Eskildsen
Part IB: How to Think About Technological Change
Illustrations courtesy of Simon Hørup Eskildsen
Part IB: How to Think About Technological Change
Illustrations courtesy of Simon Hørup Eskildsen
Part IB: How to Think About Technological Change
The Law of Accelerating Change:��The rate of change in a wide variety of evolutionary systems �(including but not limited to the growth of technologies) �tends to increase exponentially.��- Ray Kurzweil, computer scientist
Part IB: How to Think About Technological Change
This law helps us see AI as a natural progression in the rapid, evolutionary-like development �of technology.
It’s helpful to think of technology somewhat more organically.
It shows how society has continuously adapted to such advances throughout history.
Source: SingularityHub
Part IB: How to Think About Technological Change
What has occured
Part IB: How to Think About Technological Change
What might occur
Part IB: How to Think About Technological Change
TL;DR:
Part IB: How to Think About Technological Change
If you’re still here, thank you.
🫠
Next, let’s review how to think �about AI and the creative process.
Incredible Celestial Seasonings art, better than any AI hallucination
Part IC: The Creative Process
Part IC: The Creative Process
The following pages are pulled from an interview with designer Charlota Kolar Blunarova, via Off Menu.
Her insights provide a clear path for how I try to think about AI within the creative process, for right now.
Part IC: The Creative Process
Part IC: The Creative Process
Part IC: The Creative Process
Part IC: The Creative Process
Charlota illustrates a smart way to think �about AI today for creatives;
Said another way:
Stock or AI?
Part IC: The Creative Process
Next, for the Creatives reading this, I want to redefine AI…
Part IC: The Creative Process
Just like Apple relabeled Artificial Intelligence to be:
Part IC: The Creative Process
Part IC: The Creative Process
Note:
I liked this TechCrunch headline, and article, on Apple’s AI roll-out:
It’s not full of overblown optimism, it’s decidedly pedestrian, even boring.
And that’s a good thing.
Part IC: The Creative Process
The Cupertino tech giant is carefully rolling out AI where it believes it could be useful.
That means the tech won’t be included where it could be much of a threat �to the carefully crafted consumer experience of using an Apple device.
…Apple Intelligence is coming to everyday apps and features, with additions like writing help and proofreading tools, AI summaries and transcripts, prioritized notifications, smart replies, better search, photo editing, and a version of “Do Not Disturb” that automatically understands what important messages need to come through, among other things.
Part IC: The Creative Process
So, I wanted to personally redefine AI colloquially as:
Part IC: The Creative Process
So, I wanted to personally redefine AI colloquially as:
Contextual Intelligence
Part IC: The Creative Process
AI is transforming the software we use,�to be more contextually intelligent.
Stock or AI?
Part IC: The Creative Process
Which is what we humans are great at.
The artist Kehinde Wiley;
Contextual Intelligence incarnate.�
Part IC: The Creative Process
Our ability to:
Part IC: The Creative Process
By being contextually intelligent, �AI can take on more creative tasks.
As Charlota said, don’t necessarily fear �the loss of those parts of your profession, �right now.
Part IC: The Creative Process
Branding and Creative are still quite �subjective and highly-contextual. ��Offloading lower-level contextual work, �such as the hours of executional work �we all do, is a good thing.
Focus on using your mind or the deeply �creative parts of your craft.
Good creative is still good creative.
(Clockwise: Ghia, Daylight, NoPlace, Topical, Soft Services)
Part II: 1 Year Ahead
1 year: What Can We Expect
“The future doesn't belong to the ones who know the most, �but to those who learn the fastest.”
1 year: What Can We Expect
It’s not about knowing the most, �it’s about learning the fastest.
The fastest way to learn �is through doing.
1 year: What Can We Expect
Simplicity should always �be the North Star.
Don’t get caught in the middle →
1 year: What Can We Expect
Take the prior page as your compass for thinking about AI-software.
Either don’t sweat it...
Or experiment extensively to discover �what works best for you and is simplest within your creative process."
1 year: What Can We Expect
I made a list of AI tools I use today. �I figured 1 year ahead, starts with right now.
1 year: What Can We Expect
Let’s walk through them, �in four main categories:
1 year: What Can We Expect
General AI Usage
1 year: What Can We Expect
General AI Usage
ChatGPT
Use Case:�Translating PDF/Legal Docs
I now review legal contracts, and lengthy PDF �through AI, first. Then, when I go to my �lawyer, or partners, we can carefully confirm �the details, saving lengthy time, and costs.��Note: �AIs are not always accurate, so it’s important �to carefully review anything to sign, or have �a professional do so.
1 year: What Can We Expect
General AI Usage
ChatGPT
Use Case:�Visualizing information��I like visualizing information. That’s easy to do now, and pretty fast. This is me recently (during �a work break) going down an NBA rabbit hole spinning up simple graphs and pie charts on league scoring leaders.
1 year: What Can We Expect
General AI Usage
ChatGPT
Use Case:�Background research (context)��I often have to talk to businesses I �don’t know a ton about.
Whether for Pattern, or Little Plains, �I often ask ChatGPT to act as a team �member and help best inform me before �I have to do my initial calls.�
1 year: What Can We Expect
General AI Usage
ChatGPT
Use Case:�Visual assistance��While Midjourney is powerful, I often use �Dall-E for quick visualizations for things I �need right away; like turning a stock image �Into a unique generative asset to use for a project.
�
1 year: What Can We Expect
General AI Usage
ChatGPT
Use Case:�HR/Admin busywork
As an entrepreneur or studio owner, the amount of administrative work you have to �is seemingly never-ending. Having help �write JDs and other posts that can take �hours, has been incredibly time-saving. �
�
1 year: What Can We Expect
General AI Usage
Perplexity
Use Case:�Multimedia context
I find Perplexity to be great for multimedia searching, as well as wikipedia-style bouncing around.
You get images, links, and nicely formatted answers that have suggested follow-ons.
�
�
1 year: What Can We Expect
General AI Usage
Perplexity
Use Case:�Multimedia context
I find Perplexity to be great for multimedia searching, as well as wikipedia-style bouncing around.
You get images, links, and nicely formatted answers that have suggested follow-ons.
�
�
1 year: What Can We Expect
General AI Usage
Perplexity
Use Case:�Multimedia context
I find Perplexity to be great for multimedia searching, as well as wikipedia-style bouncing around.
You get images, links, and nicely formatted answers that have suggested follow-ons.
�
�
1 year: What Can We Expect
General AI Usage
Claude
Use Case:�Complex SEO optimization
�I’m not a Claude power-user, yet. ��I have it on my summer list to explore some of the deeper tasks, such as this SEO optimization example by Julian Goldie, featured on Robin Delta’s X.�
1 year: What Can We Expect
1 year: What Can We Expect
Making ads
MagicBrief
1 year: What Can We Expect
Making ads
MagicBrief
1 year: What Can We Expect
Making ads
MagicBrief
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Researching top ads
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Researching top ads
Here Anum from Pattern is walking through the process of searching for good ad templates.�
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Translating into Ads
Here Anum is looking at the information �for a selected ad to reference.
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Making briefs
Here’s she’s showing existing templates for ads �that Pattern is running through MagicBrief.
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Writing briefs
Here you can see a template for one of our �brand’s ads, as well as the references �selected being pulled in.
1 year: What Can We Expect
Making ads
MagicBrief
Use Case:�Briefs
Magic Brief also pulls in your existing ads to see �how they are performing, based on their �proprietary ranking setup and UI.
Note: I covered that information in grey here…
1 year: What Can We Expect
Making ads
Hoppy Copy
1 year: What Can We Expect
Making ads
Hoppy Copy
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Writing a newsletter
Here I’m illustrating the process of choosing an �email template to work off of, and populating the starter information based off what it offers up.
I can then easily edit information to what I need �for the brand.
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Writing a newsletter
Here I’m illustrating the process of choosing an �email template to work off of, and populating the starter information based off what it offers up.
I can then easily edit information to what I need �for the brand.
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Creating a 15 sec ad hook
Here the Pattern creative team is using Hoppy �Copy to explore different hooks for a new �product being rolled out for one of our brands.
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Creating a 15 sec ad hook
Here the Pattern creative team is using Hoppy �Copy to explore different hooks for a new �product being rolled out for one of our brands.
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Creating a 15 sec ad hook
Here the Pattern creative team is using Hoppy �Copy to explore different hooks for a new �product being rolled out for one of our brands.
1 year: What Can We Expect
Making ads
Hoppy Copy
Use Case:�Creating a 15 sec ad hook
Here the Pattern creative team is using Hoppy �Copy to explore different hooks for a new �product being rolled out for one of our brands.
1 year: What Can We Expect
Making ads
Brand.AI
1 year: What Can We Expect
Making ads
Brand.AI
Use Case:�Brand centralization / check
With Brand.AI you can upload all brand guidelines for your business - such as copy, tone, art direction. You can then use the service to not only assist in brainstorming, ads, creative – but to act as ‘brand police’ for anything coming out.
I found this helpful for teams like JAJA, that have �lots of outputs - such as regional or localized ads �and displays, to make sure all copy and visuals �are on par, per the brands’ guidelines.
1 year: What Can We Expect
Making ads
Butter�
1 year: What Can We Expect
Making ads
Butter�
1 year: What Can We Expect
Making ads
Butter�
1 year: What Can We Expect
Making ads
Butter�
1 year: What Can We Expect
Making ads
Butter
Use Case:�Building on proven ad templates
Most ads for teams I work with are dropped into Figma for commenting, or storyboarding.
Even if they’re built as videos, screenshots are �used for sharing.
1 year: What Can We Expect
Making ads
Butter
Use Case:�Building on proven ad templates
But with Butter, you can go into their site, load in your brand assets (logos, fonts, etc.)...
1 year: What Can We Expect
Making ads
Butter
Use Case:�Building on proven ad templates
Then select an ad template you like (from existing, in market ads).
For example we chose a 15 second video ad that features UI overlays pointing out product benefits.
1 year: What Can We Expect
Making ads
Butter
Use Case:�Building on proven ad templates
Then you can click into that ad (essentially reverse-engineered and editable), to start entering �in our own information…
1 year: What Can We Expect
Making ads
Butter
Use Case:�Building on proven ad templates
And voila!
A completely new video ad, with a title �and end card, our own assets, and overlaid �product detail information.
Pretty cool.
1 year: What Can We Expect
1 year: What Can We Expect
Imagery & Content
Midjourney
Use Case:�Image generation
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney
Use Case:�Image generation
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney
Use Case:�Image generation
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney
Use Case:�Image generation
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney
Use Case:�Image generation
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Magnific
Use Case:�Image upscaling��This is the before generative AI image
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Magnific
Use Case:�Image upscaling��This is the after generative AI image
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Photoshop�Magnific
Use Case:�Image generation
These are all generated imagery �for one of our brands, Letterfolk
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Photoshop�Magnific
Use Case:�Image generation
Fully AI-generated imagery
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Photoshop�Magnific
Use Case:�Image generation
Fully AI-generated imagery
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Photoshop
Use Case:�Image expansion
1 year: What Can We Expect
Imagery & Content
Photoshop
Use Case:�Image expansion
Pretty good before and after �On an internal shoot for our brand, Yield�
Before
Credit: Nicolas Leon
1 year: What Can We Expect
Imagery & Content
Photoshop
Use Case:�Image expansion
Pretty good before and after �On an internal shoot for our brand, Yield�
After
Credit: Nicolas Leon
1 year: What Can We Expect
Imagery & Content
Midjourney�Photoshop�Magnific
Use Case:�Image generation
Using all of the aforementioned tools,�We’re now getting fully advanced, �fully AI generated for brand and ad content
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Midjourney�Photoshop�Magnific
Use Case:�Image generation
Using all of the aforementioned tools,�We’re now getting fully advanced, �fully AI generated for brand and ad content
Credit: Diego Luna
1 year: What Can We Expect
Imagery & Content
Air
1 year: What Can We Expect
Imagery & Content
Air
Use Case:�Searchable creative asset management
Air’s is a competitor to Dropbox that we, and many other creative teams, use. I think they have a smarter system for organizing your creative assets.
Air has also built out AI-enabled video and audio transcription. Your videos and audio files are automatically transcribed so you can easily search for that one UGC video from a year ago that had the perfect phrase by a creator.
1 year: What Can We Expect
Imagery & Content
Figma
1 year: What Can We Expect
Imagery & Content
Figma
Use Case:�Motion prototyping
Motion design for websites is important.�When designing for development, it’s helpful for the developer to have examples of how you want elements to animate, move, and come together. That can take time. This helps save a lot of that executional time.�
1 year: What Can We Expect
Imagery & Content
Figma
Use Case:�Motion prototyping
Motion design for websites is important.�When designing for development, it’s helpful for the developer to have examples of how you want elements to animate, move, and come together. That can take time. This helps save a lot of that executional time.�
1 year: What Can We Expect
Imagery & Content
Figma
Use Case:�Design file management
In Figma, your file layers can get extremely messy. This helps clean them up so you can have a more organized file setup, for yourself and your team.
1 year: What Can We Expect
Imagery & Content
Figma
Use Case:�Text/object organization
Similarly, save time with having text/object elements come into place right away, versus manually having to sort and organize each piece.
1 year: What Can We Expect
Imagery & Content
Figma
Use Case:�Text/object organization
Similarly, save time with having text/object elements come into place right away, versus manually having to sort and organize each piece.
1 year: What Can We Expect
Marketing
Storyclash
1 year: What Can We Expect
Marketing
Storyclash
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator identification
Storyclash is tool to automate the manual process �of sourcing influencers and creators for your brand.
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator identification
In this example, we put in Open Spaces’ �Entryway Rack as a product to see who pops up.
Immediately, a long list of relevant creators �within the space is presented.
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator sourcing
Next, I can set parameters for what I’m looking for; follower range, engagement, region, gender, etc.
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator identification
Here, I’ve honed in on one creator who looks �like a good fit for our brand.
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator identification
I can see brands that this creator has collaborated �with in the past year, how often, and what type.
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator sourcing
I can look at brands that I want to learn more on.
For example, Our Place does a lot of content creator programming. So, I can pull them up and see all the creators they have worked with –
1 year: What Can We Expect
Marketing
Storyclash
Use Case:�Content creator sourcing
Also, I can look at brands that I want to learn more on. For example, Our Place does a lot of content creator programming. I can pull them up and see all the creators they have worked with –
I can then reach out to the creators I’ve identified with an assisted, and trackable outreach tool, to monitor all conversations.
1 year: What Can We Expect
TL:DR
Over the past few months, I’ve tried to gain experience by moving fast, and finding solutions that are simple.
1 year: What Can We Expect
TL:DR
Some of the tools presented are saving myself, and our teams, real time and money, each month.
�Thats time and money that can be better re-invested into other valuable areas.
1 year: What Can We Expect
TL:DR
For the year to come, these features will become easier and more integrated into the creative software suite many of us already use.
Unfortunately, many of the AI services may not last.
It’s a Cambrian explosions of AI tools right now; make sure to bet on the index, not necessarily the stock.
Part II: 10 Years Ahead
In Conclusion:
“In an age of automation, creativity is the new literacy.”
�
10 Years: What Can We Expect
I wanted to touch on two recent thought pieces by smart folks close to AI that have come out in the past few month.
They both present interesting ways to think about where AI will take software services in the decade ahead.
10 Years: What Can We Expect
Chris Paik, a tech VC, wrote an article ‘The End of Software’ at the end of May.
10 Years: What Can We Expect
He writes:��“When the internet happened, media companies viewed it as a way to reach broader audiences and reduce their distribution costs.
But what no one saw coming was that the internet not only reduced distribution costs to zero, but it also drove the cost of creating content to zero.”
10 Years: What Can We Expect
Paik theorizes that just as the internet revolutionized media by lowering content creation costs and enabling user-generated content, leading to a flood of content and the rise of new platforms – large language models (LLMs) will similarly disrupt software creation.
10 Years: What Can We Expect
By reducing software development costs, LLMs will trigger an explosion �of software tools, diminishing the value of traditional software companies.
This shift will replace established models with numerous smaller solutions, similar to how influencers overtook media.
10 Years: What Can We Expect
Here are 5 takeaways from his essay �for brand & creative specialists using �software for the decades ahead:�
10 Years: What Can We Expect
1. Increased Accessibility:
Expect for affordable software tools will be accessible to all, leveling the playing field for small agencies and individual creatives.
10 Years: What Can We Expect
2. Customization and Innovation:
The influx of new, affordable software tools will enable creatives to personalize and innovate their branding strategies.
With more options, you can tailor solutions to specific client needs, experiment with unique features, and redefine boundaries.
10 Years: What Can We Expect
3. Rapid Prototyping and Experimentation:
Lower costs will enable quick testing and refining of concepts, leading to more effective branding campaigns.
10 Years: What Can We Expect
4. Focus on Creativity:
With more technical tasks handled by LLMs, creatives can concentrate on contextually intelligent areas like ideation and strategy
10 Years: What Can We Expect
5. New Platforms and Ecosystems:
An explosive rise of new, and cost-effective, software platforms will require creatives �to incorporate ‘continuous learning’ �as a regular part of their work schedule.
Not only will new tools emerge, new ecosystems for how to create will, too.
10 Years: What Can We Expect
Bonus: The Shadow Future
If you think about the Adjacent Possible, I have no doubt incredibly exciting new creative use-cases, tools, and work processes we can’t even totally imagine are going to emerge.
The past becomes automated, �and the future gets more creative.
The cycle continues.
10 Years: What Can We Expect
Last month, a former OpenAI (the parent company of ChatGPT) employee, Leopold Aschenbrenner, published a 165 page PDF, ‘Situational Awareness’, on his AI predictions and warnings for the decade ahead.
10 Years: What Can We Expect
He predicts:�AGI (Artificial general intelligence)�ASI (Artificial super intelligence)�Will both arrive before 2030
10 Years: What Can We Expect
10 Years: What Can We Expect
I have fun story…
10 Years: What Can We Expect
Two weeks ago, I went to drop off �some pants at the local dry cleaners, Christopher Cleaners.
The young person working the counter was a college-aged student, with a laptop full of developer/coding stickers.
10 Years: What Can We Expect
I struck up a conversation, asking if �he was into coding, and what he was studying.
He said yes, and machine learning.
10 Years: What Can We Expect
I asked if he had read Leopold’s essay?
He had.
I asked if he agreed with his prediction for the next decade?
10 Years: What Can We Expect
He said, ‘not entirely…’
Then another customer came in, �so I said thank you, and politely left.
10 Years: What Can We Expect
…But I wanted to know why, ha!
10 Years: What Can We Expect
Later that day, I went on google, got �the dry cleaner number, and called it up.
He picked up.
I said, ‘hey, I’m the pants guy that asked you about ML and Leopold’s essay…’
10 Years: What Can We Expect
He laughed and said, ‘yes!’
I asked, ‘so, why don’t you agree �with his predictions??’
10 Years: What Can We Expect
He said, ‘well, I am skeptical of the effectiveness of the government.
Issues that are supposed to be non-partisan (i.e. education, pandemic, etc.) become controversial.
If the nation can't efficiently rally behind some kind of AGI dream, I am not sure if we can build massive data centers to train AI.’
10 Years: What Can We Expect
That made sense to me.
Lastly, I asked what he thoughts of AI for normal and small businesses?
10 Years: What Can We Expect
He answered that while he loves ML, �he doesn’t see how AI directly affects his family’s dry cleaning business, �for example.
Customers still come in and do business in person, and we were talking on the business phone line of the dry cleaner.
There is just something about doing the business, in person.
10 Years: What Can We Expect
I agreed but said, in the early 2000’s one may not have imagined someone walking down the street, finding the business online, with photos, reviews, directions, hours – and calling him up – from a mobile phone, all with a few swipes.
Nor that I had just come in and paid �with just a tap of the same phone.
10 Years: What Can We Expect
He agreed to that.
And we agreed to talk more.
10 Years: What Can We Expect
I’m sharing this presentation �with him today, too :)
10 Years: What Can We Expect
So,
I think Leopold may be overestimating what technology is possible by 2027
but we may be underestimating how much the world around us will just naturally change because of �technology over the coming decade…
10 Years: What Can We Expect
Thanks to Ryan Zhang, from Christopher Cleaners for the conversation, and permission to share.
If you happen to be in the West Village, interested in machine learning, and have some clothes that could use sprucing up – stop by and say hi :)
10 Years: What Can We Expect
Lastly, here are his follow up thoughts �on how AI does, or doesn’t, affect dry cleaning.
It’s thoughtful and I wanted to share �as he wrote it:
In Conclusion:
In Conclusion:
TL;DR:
For Building and Creativity in the Age of AI
Thank you.
Emmett Shine��X: @emmettshine�IG: @emmett�LinkedIn: emmettshine��Web:�emmettshine.com�patternbrands.com�littleplains.co��✌️