Lars Hansson

Founder and LAH, Former Head of Design @RGA

[Lars Hansson @LarsHansson]

Excited to launch https://t.co/Lic6LllhrW!

We empower creative, strategy, and marketing teams in adapting to generative AI by augmenting workflows, creating human-centric AI solutions, and innovating in products and services.

I'm seeking collaborators to form a network of like-minded professionals, enabling us to scale and provide solutions of varying complexity. DMs are open.

Check out our website (link in replies).

[Chris Dixon @cdixon]

I recently went on Rick Rubin’s podcast.

We got to talk about:

- internet history
- little tech vs big tech
- internet economic models
- composability & open source software
- blockchains & tokens
- software as art

And much more.

You can listen here:

https://t.co/jiVje7UCcm

[Lars Hansson @LarsHansson]

@Timb03 Automate to send a survey that filter out genuine interest + an opportunity to validate the lead before the call.

[Lars Hansson @LarsHansson]

@thatroblennon Happy that Google AI services are advancing, but their naming is not making it easy for users to understand...

The Gemini family: "Nano, Pro and Ultra".

But, you need "Advanced" to access "Ultra"?

Why not just call them "Gemini Pro" (Free) and "Gemini Ultra" ($19.99)…

[Lars Hansson @LarsHansson]

The age of static branding is over. AI is enabling dynamic, evolving brand experiences.

Marketing agencies are now leveraging generative AI to craft brands that aren't just sophisticated but deeply authentic, marking a pivotal shift in branding strategies.

Forrester Predictions 2024 highlights,

“In 2024, agencies will invest heavily to build bespoke, enterprise-ready AI solutions, formed by a combination of foundational AI models from tech partnerships, their own audience and creative intelligence, clients’ first-party data, and the brand’s standards.”

This underscores the movement from traditional branding to dynamic AI-driven tools, enabling brands to more effectively adapt to their target audiences with experiences that are not just relevant, but deeply engaging.

The focus now shifts towards creating adaptive, responsive brand experiences that evolve with customer interactions and preferences.

However, integrating AI in branding brings challenges, particularly in expressing a brand’s nuanced characteristics. Mastering this complexity is crucial for authentically representing a brand's unique qualities.

The potential of AI in transforming brand marketing is immense, marking the beginning of a new era in branding that is not a static representation, but a dynamic, interactive, and multi-faceted experience.

What are your thoughts on the future of AI in branding?

[Lars Hansson @LarsHansson]

I do think Google will be a dominant player in AI, but I wonder how long it will take…

It seems like they are too big and fragmented to quickly develop new products or implement them in a meaningful way into their suite.

NotebookLM is the most interesting product I've seen so far, and that seems to have come from a small team in Google Labs.

[Lars Hansson @LarsHansson]

The website has some interesting content about the Large Action Model (LAM). I can see how AI, in general, could become really good at learning how to navigate UIs and perform actions.

https://t.co/4z0eekJ6L5

[Lars Hansson @LarsHansson]

Interesting device! Love that Teenage Engineering is doing the design but another gadget in the everyday carry could be a barrier…

The video shows some interesting details about the interaction with apps. Could this be the next iteration of the mobile interface? Siri 3.0?

[rabbit inc. @rabbit_hmi]

Introducing r1. Watch the keynote.

Order now: https://t.co/R3sOtVWoJ5

#CES2024

[Yann LeCun @ylecun]

Genuine question about image generation:
If someone uses a generative AI tool to produce an image that is substantially similar to a copyrighted piece (a drawing, painting, movie screenshot, etc), who should be liable for copyright infringement?
Should it be:
A. the company producing the tool?
B. the (human) creator of the piece?
C. the person or entity posting/publishing the piece?
D. the communication platform through which the piece is distributed?

1. Copyright law protects against unauthorized exact or near-exact copies of a painting, photo, movie, or other visual piece.
2. When a person distributes a sufficiently similar copy of an art piece, it's a violation of copyright regardless of the tools and process used to produce it.
3. The liable person is the person *distributing* the piece, not the artist, and not makers of the tools.
4. Image generation systems are trained to generate images that are on the "manifold" of nice-looking images. Obviously, the training images are on this manifold.
5. Hence, sufficiently detailed prompt will produce images that are substantially similar to images from the training set. It is not at all surprising that a prompt like "The Batman movie, rooftop scene, screenshot, 4k..." will produce an image very similar to an actual screenshot from the movie.
6. Whether using publicly available-yet-copyrighted screenshots and other materials as part of the training set constitutes a violation of copyright is a separate question. As far as I can tell, this question is not legally settled in the US.

[Google Fonts @googlefonts]

New Fonts! Wavefont and Linefont by Dmitry Ivanov are non-text #variablefonts to render data. #datavis 📈 https://t.co/2zOkIrswUF

[Lars Hansson @LarsHansson]

I updated my website with new and past work.
Let me know what you think and let’s work together!
→ https://t.co/7WnECnNLux

[Lars Hansson @LarsHansson]

Exploring #midjourney
[vibe, subject and objects] to create mood-boards and brand assets

#AIart #AIArtwork #AIArtCommuity

[Lars Hansson @LarsHansson]

Now that generative AI is going mainstream through tools like Google Docs, Adobe, Canva and others, I'm noticing some recurring challenges that come up when offering solutions using AI to clients.

1. Pricing
Many clients don't fully understand AI's capabilities yet, leading to unrealistic expectations around speed and cost. In our industry, we’re accustomed to hourly billing, so the assumption is AI should make everything faster and cheaper.

But the reality is AI can provide immense value far beyond efficiency gains. While AI reduces production costs, it also enables things like: more customized experiences that increase customer loyalty, faster adaptation of products and services to market changes, achieving greater output with smaller teams.

I believe we need to shift to value-based pricing that factors in these benefits. The challenge is quantifying that added value when the technology remains emerging with limited proof points.

2. Copyright
Clients want to own and be able to copyright their content. But in reality, regulations around copyrighting AI content vary widely internationally. For example, AI-created works currently lack protection under U.S. copyright laws.

There has been an increase in solutions that involve training customized AI models on proprietary data with clear ownership permissions. While this approach has mainly targeted large enterprise clients so far, advances in open source models are making it more accessible for businesses of all sizes. Additionally, some services like Shutterstock guarantee no copyright infringement and provide compensation to creators whose work trains the AI.

3. Legal Uncertainty
There’s a lot of uncertainty surrounding the legal risks for companies providing AI services since regulations haven't kept pace with the technology's rapid development, leaving clients in an ambiguous position lacking accountability.

However, the industry is responding by offering protections. Companies like Adobe covers legal costs for copyright lawsuits regarding its AI tools. Groups like the Partnership on AI are also advancing best practices around issues like bias and transparency. But regulatory clarity remains minimal.

How are you navigating these issues around pricing, copyright, and legal uncertainty when it comes to selling generative AI services? I'm curious to hear what obstacles others are facing as this technology continues to develop.

[Lars Hansson @LarsHansson]

@camdiamond Maybe you don’t need an engineer? #NoCode

[Lars Hansson @LarsHansson]

@asaquiros @ciguleva Inspiring! I haven’t used Processing in a while and get intrigued to pick it up again now when we have help from ai to write code.

[Lars Hansson @LarsHansson]

@hanne_van_briel Very nice!

[Lars Hansson @LarsHansson]

Yes, I’ve been thinking a lot about fine tuning and this seem to be a really good way to do it!

[Pietro Schirano @skirano]

Ever wondered how to fine-tune SDXL for any style you desire?

I've got the simplest guide for you using @replicatehq.

Let's dive in!

[Lars Hansson @LarsHansson]

@hibrunocafe Thanks Bruno 😎

[Lars Hansson @LarsHansson]

Exploring #midjourney [vibe, subject and objects] to create mood-boards and brand assets

#AIart #AIArtwork #AIArtCommuity

[Lars Hansson @LarsHansson]

@cj_zZZz Great list, thanks for sharing!