cobalt.

Founder

Google

Technical Consultancy for global retail brands, driving increased conversions and profitability within mobile ecommerce.

Technical Specialist, Mobile Ecommerce

Working with teams in Local Search, Shopping, Geocoding.

Program Manager, Geo

Dotted line manager for Google Earth VR vendor team.

Vendor Manager, Daydream - Earth VR

Webmaster for Street View and Geo Imagery related sites. Specific interests in Virtual Reality and spherical photography.

Lead Web Developer, Street View

Bringing rapid prototyping, contextual animation and material design to product front-end.

UX Prototyper, Travel

Webmaster on numerous $M+ budget marketing campaigns across the globe.

Webmaster

AWE (Augmented World Expo)

Master of Ceremonies for the main stage of Europe's largest Augmented Reality Exposition with an audience of 1,000 attendees.

Conference Host

Musemio

Technical advice on UX guidance and product excellence for a mobile app that allows children to explore history and arts through gamified mobile VR quests.

Advisor

Immersionn

Advisory Board

meepl ⎜ Fision AG

Specialist technical advice and partnership development within the fashion and AR/VR industries. Company successfully acquired by Zalando.

United Nations Office at Geneva

Summit design and delivery outlining the potential uses of Augmented Reality to amplify the UN's external communication strategy.

Advisor – Augmented Reality

TEDxZurich

Always looking for English language speakers to present at TEDxZurich each November.

Speaker Liaison

Edinburgh Interactive

Assistant Organiser

RiverCultures.tv

Owner

Google Technical Specialist | Prompt Engineer | Vogue Photographer | Metaverse Tour Guide – AI Art, VR/AR, Fashion, Architecture

After 16 years at Google, I founded the technology consultancy Cobalt. My team helps companies save money and reduce employee burnout by assisting them in integrating AI into their workflow. Advisory partners include Alexander McQueen, Zaha Hadid Architects, Perkins Eastman, and Curiosity Studios.

I deliver keynotes on subjects including Generative AI, The Metaverse, Virtual Reality, Wearable Tech, and Virtual Beings. I also provide executive briefings, moderate panels, and guest on podcasts.

At Google, I made impactful contributions across a wide range of teams and partner relationships and invested in office culture, and new hire onboarding strategies. Roles include Product Manager, Program Manager, Solutions Consultant, and Frontend Engineer.

I actively build and mentor AI and Metaverse communities. With over 11,000 LinkedIn followers, I’m a trusted voice in AI and listed as one of the Top 100 AI Artists. I have been interviewed by Vogue, The New York Times, and The Wall Street Journal.

As an early-stage startup advisor, I help companies articulate their value proposition and optimize user onboarding. Recent exit: virtual try-on app meepl acquired by Zalando for $XXM. I have an extensive technology and business network built across 30 years in the industry.

Congratulations to Kevin Gim and Joohyung Kim of Boston Dynamics and KIMLAB (Kinetic Intelligent Machine LAB), who assure us that the future of robotics isn't entirely humanoid. Their Ringbot: Monocycle Robot With Legs is a beguiling mix of the motorized Wheelitis of the 1930s and The Revenge of the Sith's Tsmeu-6 wheel bike.

Credit to Udio for the music track, a tool which continues to amaze and delight me.

#ringbot #kimlab #robotics #monocycle #locomotion #bostondynamics

---

https://lnkd.in/ej6C-Gr8

"Ringbot aims to provide versatile mobility by replacing the driver and driving components of a conventional monocycle vehicle with legs mounted on compact driving modules inside the wheel. The Ringbot prototype features a wheel and two driving modules located inside, each equipped with a 3-DoF leg for balancing, steering, and legged motions to assist monocycle driving. The driving control is achieved through decoupled speed controller and steering controller. In addition, active-legged motions are implemented and managed through a finite-state machine. The controllers for wheeled driving and legged motions were tested in a simulation environment, as well as on the hardware prototype, to verify the concept of a monocycle with legs and evaluate the prototype's capabilities."

Rupert Breheny

What is sentience? How do we define intelligence and self-awareness? How do we know if our friends and family are merely "psychic zombies" or, in fact, possess full consciousness? Do animals have a capacity for theory of mind? Would aliens? Will robots?

For centuries, these questions have been an academic discourse, but with recent advances in AI, society will need to start wrestling with them on a larger stage.

Some will ascribe human motivations and sensibilities to AI, especially when embodied in robot form. There will be calls for android rights in a replay of the civil rights moment of the 1960's.

I don't believe sentience will result from merely expanding token capacity for large language models, even though interacting with such systems will feel like a conversation with an intelligent and self-aware intellect. There is a danger that they will become so good at faking emotional states that they may inspire us to act against our own interests, manipulated as we are by the ultimate sociopath.

There is a lot of reading to be done here; both David Chalmers and Yann LeCun speak eloquently on the subject, and the movies A.I., Her, Ex Machina, and The Creator all ask intriguing questions about how humans will choose to view and interact with artificial intelligence in the very near future.

Now is the time to start having these conversations in earnest so we aren't caught off guard by imminent changes in circumstance.

#midjourney #aiart #rupertbrehenyart #consciousness #ai #sentience

From Runway to Pika and soon Sora and Lumiere, we're spoilt for choice in the text-to-video field. But nobody provides a single tool that can take you from script to ship in a single interface. That might be about to change with the imminent rollout of Lightricks' LTX Studio.

Until now, an AI video production workflow might have started in Word, routed through ChatGPT, diverged through Midjourney, Runway, and Eleven Labs, and come back together in CapCut. That's a lot of moving parts and a long round trip if you need to change a creative decision at an earlier point in the process. This level of convolution takes time, energy, and a raft of different subscription fees. 

LTX Studio simplifies a lot of that with an all-in-one package. Since the start of my Midjourney experiments, I've been using their Motionleap app to animate images for over a year, and I have enjoyed their tools and interface.

Their initial visuals aren't on a level with Sora yet, but I expect them to swap out foundation models as advances are made, so expect steady progress.

What interested me most at their launch event was that they gave the stage to Ari Folman, the Golden Globe-winning director of The Congress, a movie that predicted the future of film production over 10 years ago. He highlighted a few key features of the rapidly emerging AI technology that will drive the film studio of the future.

1) It could reduce the production time of an animated movie by up to three years

2) It can prevent burnout of an overworked technical staff, running to brutal deadlines

3) It massively de-risks developing new content by slashing budgets and timescales

This is all music to a movie studio's ears, but maybe it won't be so well received by debt-laden Film School graduates or animation tween artists working their way up through the industry.

On the plus side, the competition will no longer be decided purely by your technical ability behind a camera or in the edit suite, but by the calibre of your ideas and your ability to tell stories that fascinate and move an audience. Now is the time to dust off those film treatments and develop them into a full-length film yourself. This is an undoubted win for creativity, and an exciting future lies ahead for the industry at large with this influx of new talent and fresh stories.

Lightricks will need to stay nimble here. Runway is accelerating, and OpenAI's Sam Altman is courting Hollywood. The teased Premiere Pro AI integration must be progressing rapidly, a major boost to a toolset many professional filmmakers already use daily. 

This will be a competitive space, and creators have never had more powerful tools to help them share their vision on the most accessible platform that has ever existed. So, let's get to it.  

Quiet on the Set: Lights, Cameras, Action!

#ltxstudio #runway #premierepro #lightricks #openai #sora #generativeai #filmmaking #vfx #hollywood #moviestudio #storyboard

I have been fascinated by technology my entire life. I am mesmerized by the magic that happens when advances in science are transformed into objects and tools that either open up whole new worlds of possibility or just make our days a little more comfortable.

I've spent the last 25 years working in various technology fields, and as much as I would like to convince you otherwise, it's all been pretty much accidental.

What I lacked in career planning and ambition, I made up for in curiosity and enthusiasm; both traits have served me well. I read books and articles, cultivate a broad view across various industries, take long walks, and think about things. I try to stay ahead of the pack with an eye towards the future with a focus on the human side.

Sometimes, I get it right: I made my first website 28 years ago when I knew everybody would one day have a web presence. Sometimes I'm a little early: I bet big on XR when I built Google's first dedicated VR lab 8 years ago in anticipation of a wave of broad headset adoption. Sometimes, I get it wrong: light field cameras have had no discernible impact on consumer photography. Sometimes, I even make decent money: I blew my entire annual bonus on Tesla stock after my first test drive.

One thing is clear; the rapidly increasing rate of change. Compounding advances in a range of deep technologies are occurring in multiple industries at once, all made even more impactful with the onset of AI. I love that I was born in an age where I get front-row seats to see the world realign in real time. 

I love talking about it, too. My core leadership skill at Google was my ability to communicate complex ideas clearly in a way that allowed lay people to participate. I relish sharing these insights and am gratified that people value my stories about possible futures. I want to thank everybody who allowed me to share some of them in speeches and briefings around the world.

I aspire to become a better communicator, and I deeply appreciate all you kindred spirits who do the same on your stages, podcasts, YouTube channels, and Medium articles.

Aragorn Meulendijks, Colin Fitzpatrick, Dale P., Francesca Rosella, Irina Raicu, PhD, Marco Morales, Marily Nika, Ph.D, Mike Chatterton, Thomas Rothery

We move forward together and bring as many on the journey as are ready to take their first steps. Here's to the future …

#future #ai #technology #speaking #advisors

I'm trying out Udio today, and I'm impressed. This is the latest among several text-to-music AI tools, and I've been getting output that I would actually want to sit back and listen to, and I can't say that about many of them so far. Congratulations to the team over there, a stalwart cadre that includes some exGoogle DeepMind engineers doing for music what Midjourney did for imagery.

David (Fengning) Ding, Yaroslav Ganin, Justin K. Chen, Andrew V. Sanchez

Here, I'm illustrating a story I've been discussing with my friend Aaron Shon for a while now. Exploring the bittersweet knowledge that you embark on a journey that you won't live long enough to see to completion. But then, that's true of any gardener or cathedral builder of ambition. Some endeavours are bigger than any one human, and a trip to colonise other solar systems certainly ranks amongst them.

I'm still not happy with the glitching and loss of clarity I get from Runway or Pika animations, so I continue to use my favourite depth map extraction tool, LeiaPix, and then spice them up with some CapCut effects. It may be more accurate to call this an animatic than a trailer, but I like the balance of motion to image integrity.

#udio #midjourney #leaipix #capcut #space #trailer #rupertbrehenyart

There are thin places, specific locations where the veil between this world and the next is at its slightest, and the attuned spirit can sense that land which lies beyond.

Treasure such a place if you find one, and share it only with those you trust beyond the bounds of life itself. These are dangerous paths, and to cross such a threshold imperils the uninitiated with the opium of amnesia, which obscures the path of return.

#midjourney #aiart #rupertbrehenyart

It's going to be a hot, hot summer. When they talk about us moving into an accelerating phase, they aren't kidding:

OpenAI, Google, and Mistral have released new AI models within 12 hours. The fact that both GPT-4 Turbo and Gemini Pro 1.5 are 'multimodal' systems will open up many new use cases; now, we can use images, audio, and video in our prompts.

I recently listened to the Lex Fridman Podcast with Yann LeCun (Meta's Chief AI Scientist). He's quite unimpressed by the current LLMs' ability to understand the world. So, his take on researchers focusing on 'objective-driven' AI with the ability to reason and plan to bring about better world understanding is intriguing.

He uses the colorful metaphor of a child locked in a closed, windowless room with a giant pile of books who is somehow expected to make sense of the world. The reality is that children need to run outside and see things with their own eyes, play with sticks, climb trees, and so on. You learn about the world by seeing and doing, not just reading. 

Embodied AI (ubiquitous robots) are going to be a huge boost here.

#openai #google #meta #mistral #robotics #ai #llm

---

“We hear a lot of people saying: ‘Oh my God, we’re going to get [artificial general intelligence] within the next year,” said Meta’s chief AI scientist, Yann LeCun, responding to a claim from the xAI founder, Elon Musk. “It’s just not happening. We have AI systems that can pass the bar exam, but they can’t clear up your dinner table and fill up the dishwasher. We have systems that manipulate language, and fool us into thinking that they are smart, but cannot understand the world.”

theguardian.com • 3 min read

AI race heats up as OpenAI, Google and Mistral release new models

Sometimes, I'm blessed with highly detailed dreams with great visuals and a degree of rational thought while I'm having them. This unusual rendering is a pretty accurate portrayal of what I saw last night. It sounds odd, given the atmosphere, but this was very clearly set in Amsterdam. I guess this must be New Amsterdam circa 2089.

#midjourney #aiart #rupertbrehenyart

As excited as I am by AI, I often read contrarian views to ensure I don't fall too far down the rabbit hole. 

People perceive chat-based Large Language Models (LLMs) as intelligent despite their lack of true reasoning. Users often accept the first response or repeat prompts until they get the desired result without fully engaging.

The are some potential adverse outcomes in accepting the output of these systems at face value:

1) Perpetuating misinformation as users accept LLM outputs as fact.
2) Undermining trust in technology as LLM limitations become apparent.
3) Ethical concerns around transparency and potential deception.

The industry's efforts to make LLMs sound more human through techniques like Reinforcement Learning, allow these systems to sound smarter than they really are.

An honest appraisal of their limitations is crucial to ensuring the responsible development and deployment of these technologies.

I appreciate that Baldur Bjarnason has a book to sell, but I also resonate with his thoughtful account of the failings of current LLM responses.

https://lnkd.in/eAb5CWHy

#llm #ai #chatbot #misinformation #hallucination

The smell of dogwood always brought her right back to those moments. Those long summer weekend afternoons in the garden of the house she grew up in. The shelter of this, her favourite tree, scant protection in the rain but the perfect canopy against the sun.

A refuge from the chaos of the house as aunts, uncles, nieces and nephews came and went, wondering why she was so content to sit and read all day.

She'd tried being "like the other kids". It was overrated. 

#midjourney #aiart #rupertbrehenyart

The Tiger of Forli was as beautiful in deathly repose as she was fearsome in the heat of battle—if you allow me a little creative licence … 

We all know Joan of Arc, but Caterina Sforza was no Damsel in distress, either. Despite her illegitimate birth, she was afforded a comprehensive education and became an influential and powerful figure in Renaissance Italy. 

An early marriage was not to last, and after her husband's assassination, she became a shrewd leader in her own right. Known for her military prowess and political cunning, she defended the fortress of Forli against the forces of Cesare Borgia and allegedly flipped off her attackers from the castle walls.

Sforza was a force to be reckoned with in both battle and defeat. On her eventual capture, she was permitted to live as a guest in Florence, and devoted herself to her children and a series of alchemical experiments, which included an alleged attempt to kill the Pope with a poisoned letter. 

Amassing many lovers, husbands and children, Caterina outlived many of her rivals, the Borgias included and managed to maintain control of her territories until her death in 1509 at the ripe old age of 46.

#midjourney #aiart #rupertbrehenyart #damsel #italy #history

Join me to talk Art, Fashion and AI in 30 mins (16:40 CET) on Day 3 of YIQIN ZHU's Fashion Metaverse Conference:
https://lnkd.in/gUKTRybH

Thanks to a kick-ass creative panel:
Aditya Mani, Jens LAUGESEN, Christopher Sicurella, Wildy Martinez, Micael Carvalho

Available via video Streaming:
https://lnkd.in/gRQJ4YwR

Or dive into the metaverse with Spatial:
https://lnkd.in/g4JgmXDe

Unveil the unseen world! Capture fleeting spirits and luminous auras with Dr. Mesmer's Aethereal Stereographer, a true marvel of the modern age. The very Aether itself becomes visible for spiritual vision and enlightened study, revealing the mystical dimensions that lie beyond the veil.

Do shadows whisper secrets?

Mesmer's Marvel reveals the phantoms lurking at the edge of sight. For those who dare, document the spectral realm and commune with the spirits of the beyond.

This gateway to worlds unimagined will surely delight and illuminate discerning gentlemen and ladies of curiosity, whether they be dedicated scholars of the occult sciences or adventurous amateurs in the realms of the fantastic.

The Aetheric Research Society - Expanding Metaphysical Knowledge Since 1872.

#midjourney #aiart #rupertbrehenyart

Remember the good old days when deep fakes needed Gigabytes of Tom Cruise movies as training data and days of model training? Well Microsoft just revealed their VASA-1 project, which focuses on generating lifelike audio-driven talking faces in near real-time from a single reference image.

The system uses deep learning models to animate a virtual character's face to match the audio input, creating a realistic and synchronized talking animation. This tech has applications in areas like virtual assistants, video games, film production, and election interference!

Like I keep telling people, we need to sit down and have a serious talk about not taking social media videos at face value. The post-truth world is well upon us, so let's keep our wits about us and commit to only using this for good.

Congratulations to Jiaolong Yang and his team at Microsoft Research.

https://lnkd.in/eDvGVF8w

"We introduce VASA, a framework for generating lifelike talking faces of virtual characters with appealing visual affective skills (VAS), given a single static image and a speech audio clip. Our premiere model, VASA-1, is capable of not only producing lip movements that are exquisitely synchronized with the audio, but also capturing a large spectrum of facial nuances and natural head motions that contribute to the perception of authenticity and liveliness."

#microsoftresearch #chatbot #faceswap #deepfake #ai

Owlcats are a classic good luck charm and are known to ward off incursions of both snarffles and boglets.

In these times of recurring supernatural infestation, a small felt totem such as this tied to your bed frame can ensure a good night's sleep untroubled by the call of restless fey.

#midjourney #aiart #rupertbrehenyart

If, while lost in thought, you happen upon a fairie lane, be sure to tread carefully. Even if you are not lucky enough to meet one of the fey, you may still be permitted to enjoy their immaculately tended ways and gardens.

#midjourney #aiart #rupertbrehenyart #dioramatuesday

Sold at over a billion dollars apiece, the Goliath unit was deemed the ultimate deterrent. Atomic power permitted a century of service without refuelling. Magnetically accelerated ordinance drove a withering hail of 1,000 flechettes per minute. Mechs were to be the sixth and final dimension of warfare.

They inspired patriotic zeal amongst the volunteers who underwent the medical extraction procedure, allowing the pilots to present the smallest possible target encased at the armoured core.

The marketing department was instructed to avoid depictions of "brains in jars" and instead pitched the wish fulfilment of the ultimate full-body upgrade for those granted the honour of protecting the homeland. 

But freedom from the necessity of food, sleep, or physical discomfort came with the loss of endocrine modulation and, ultimately, the buildup of toxins. The pilots, in their fever dreams, imagined enemies all around and took a terrible toll on their own before they could even leave the parade ground.

A billion dollars apiece, and broken down for scrap at a dollar a day on the sub-continent. And as the inner cores were finally breached, a blessed release in the baking equatorial sun.

#midjourney #aiart #rupertbrehenyart

At last, some adventurous new casting decisions for this Summer's imminent franchise reboots. 

“Sometimes the only way to stay sane is to go a little crazy.”
 Melissa McCarthy is Harley Quinn

"Why do my eyes hurt?"
Will Smith is Neo Anderson

"Toto, I've a feeling we're not in Kansas anymore."
Zendaya is Dorothy Gale

"Someday a chocolate rain will come and wash all this fudge off the streets."
Travis Bickle is Willy Wonka

I've been playing around with Musavir.AI's new Avatar function and I'm pretty impressed. Just click the little toolbox icon top right of the prompt box and then upload your own headshot into the MyAvatar dialog. I compare this approach to the output of Midjourney's --cref parameter using the same reference image and prompt, and its a clear win for Musavir.

1–4 are Musavir
5–8 are Midjourney

#musavir #midjourney #aiart #rupertbrehenyart #movie #casting

I know what you mean, Myke. When I first landed on this benighted rock, they all looked the same to me, too. Apparently, they can tell each other apart by tiny differences in the pigmentation of their dermis and ocular fringing.

I should have paid more attention to their headwear at first. That's the biggest pointer. It turns out it's not even clothing; it's attached to their head stalks. It can be all different colors and lengths, and they seem to spend many of their wake hours poking at it. These humans are crazy!

#midjourney #aiart #rupertbrehenyart

Cool air on the face, as dappled light dances across eyes half-closed in an ecstasy of the senses. The snap of a twig underfoot counterpoints the swish of shed leaves swept along on each languorous footfall. The smell of good earth contrasting the oily-green tang of fresh bluebells at the back of the nose.

#midjourney #aiart #rupertbrehenyart

Taplio

Rupert Breheny's Linkedin Analytics

Rupert Breheny

What is Rupert talking about?

Rupert Breheny's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Rupert Breheny's Linkedin Analytics

Rupert Breheny

What is Rupert talking about?

Rupert Breheny's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Rupert Breheny's Linkedin Analytics

Rupert Breheny's Best Posts (last 30 days)