OpenAI to the NYT show us your reporters notes we want to see if its even copyrightable.
NYT responds, “given the wildly improper scope of this request, one has to wonder if a chilling effect is exactly what OpenAI, who appears to have stolen from millions of content creators, is hoping for.” https://www.documentcloud.org/documents/24791465-nyt-openai-letter
AI, Synthetic Data & Media
Transcript: Robo DJ — YouTube invests in AI-generated music (ft) “So YouTube is trying to make it bigger and make it a proper product. It’s about kind of enticing people to Shorts and making them want to stay there and watch videos there and make videos for Shorts. They have recently kind of come to these big music coming in and said, what if we just pay you, you know, a lump sum of money in exchange for bringing more of your artist into this and allowing more of their music to be used to train our AI. ” (“lump sum” -ed)
Instagram’s ‘Made with AI’ label swapped out for ‘AI info’ after photographers’ complaints (verge) “Made with AI” label to “AI info”. While YouTube has added reporting and context feedback
AI’s Brain Fog Won’t Stop a Reckoning for the Arts (bloomberg) It will take a few more years for banks and health-care firms to solve AI’s hallucination problem, but creative industries already face some harsh disruptions. (My mantra the last two years -ed)
Brazil data regulator bans Meta from mining data to train AI models (apnews) The decision stems from “the imminent risk of serious and irreparable or difficult-to-repair damage to the fundamental rights of the affected data subjects,” the agency said in the nation’s official gazette.
Why GPS Is Under Attack (nyt) The American GPS network that was once the gold standard is at risk of becoming a relic as Chinese, Russian and European systems modernize.
AI washing: Silicon Valley’s big new lie (mikeelgen) The cumulative effect of AI washing is that it leads both the public and the technology industry astray. It fuels the delusion that AI can do things it cannot do. It makes people think AI is some kind of all-purpose solution to every problem — or a slippery slope into dystopia, depending on one’s worldview.
SoftBank to prioritise AI deals over share buybacks despite pressure from Elliott (ft) – Ed note: 2019 SoftBank owned 4.9% of Nvidia and sold it all for a $3.3B profit (todays value would have been 160B)
Provenance
The Fairchain platform will be closing, effective August 1. (fairchain) All records and documentation stored on Fairchain will remain accessible for existing clients until this date, including certificates of title for artworks and sales contracts, the terms of which will continue to be upheld. Effective immediately, no new accounts will be able to be created on the platform. (yada yada yada block chain artists yada- ed)
Video Editing Tools
Runway Gen-3 Alpha: text-to-video model
pikimov.com This non AI video editor is open, web based and never once asked me to login. I am in shock. This is the lesson for every startup out there – every time users hit a speed bump they hit next and forget your product for good. I shared, used and bookmarked this for its utility + their powerful demonstration and understanding of UI + UX
CapCut Free all-in-one video editor for everyone to create anything anywhere
Privacy
I was served an Ad that featured an AI Photo of myself on Snapchat. What can I do? (reddit) Snapchat pulls all your photos once you allow for it to access them and then uses one you did not post to share with third parties for advertising. The only thing here I see changing is them targeting his friends and not him in the future. (something something COPPA, kids, something , just wtf -ed)
Cloudflare rolls out feature for blocking AI companies’ web scrapers (cloudefare)
Forget privacy, young internet users want to be tracked (ft) This is not because they are blind to the importance of online privacy. It is because they are realistic about the privacy that is available. They know that if you own a smartphone and don’t want to disable useful things like maps then your location is already being watched.
OpenAI updates its ChatGPT macOS app to encrypt locally stored conversations, after a user discovered that the app was storing chats in plain text (threads) OpenAI = untrusted -ed. Seems like their internal coms were owned by a cracker and they were mum (nyt)
Ticketmaster at this point has a standing, un-dated we have been owned statement online = evergreen.
Lexicon
“self-inflated feedback loops”
The AI we could have had (ft) But the rebels at the lab thought this kind of automation was the antithesis of true responsiveness. They saw human relations, art and identity as open-ended, always-evolving ecologies that could not be reduced to the thermostat’s simplistic model of optimisation. Can one really pinpoint the “right” cinema, music or loved one in the same way as the right temperature of a room? Today’s TV, music and dating apps seem to think so. The Boston contrarians did not.
Gen AI: too much spend, too little benefit? (goldmansachs) “Given the focus and architecture of generative AI technology today… truly transformative changes won’t happen quickly and few—if any—will likely occur within the next 10 years.”
The Voices of A.I. Are Telling Us a Lot (nyt) Even as the technology advances, stubborn stereotypes about women are re-encoded again and again – When Sky disappeared, ChatGPT users took to the company’s forums to complain. Some bristled at their chatbots defaulting to Juniper, who sounded to them like a “librarian” or a “Kindergarten teacher” — a feminine voice that conformed to the wrong gender stereotypes. They wanted to dial up a new woman with a different personality. As one user put it: “We need another female.”
EvTexture: Event-driven Texture Enhancement for Video Super-Resolution (github)This allows for gradual refinement of texture regions across multiple iterations, leading to more accurate and rich high-resolut ion details
Mustafa Suleyman’s statements about copyright and AI (stackdiary) He noted that content on the open web has traditionally been considered fair use, allowing anyone to copy, recreate, or reproduce it unless explicitly restricted.
Quora-owned AI chatbot platform Poe (wired) is providing users with downloadable HTML files of paywalled articles from outlets including NYT, Forbes, and The Atlantic
Perplexity’s grand theft AI But Perplexity has taken it a step further with its Pages product, which creates a summary “report” based on those primary sources. It’s not just quoting a sentence or two to directly answer a user’s question — it’s creating an entire aggregated article, and it’s accurate in the sense that it is actively plagiarizing the sources it uses.
“And when I opened the curtains they were taking the set away and packing up for the day, the cameras and lights turned off. ;
The darkness replaced with strip lights and the grey skies, the blind whirring of machinery.
I’d like to write a beautiful story about love.” Stanley Donwood
Storytelling is a fundamental human need, an art that is a moral imperative, the mediums through which we express this need have changed from spoken word passed down through generations to AI and other technologies that are emerging now.
The future of storytelling lies in the tales we tell, and the tools we use to create to tell them. Stories shape our world and imagination.
New tools can break barriers and democratize storytelling, allowing diverse voices to inspire hope, foster empathy, and guide positive change. Tools with the potential to be as revolutionary as the alphabet or the camera, allowing for new narrative exploration and more importantly the ability for us to impact our collective imagination. Text to screen is powerful advancement in our ability to tell stories.
Tools and technology has always impacted storytelling, from silent films to interactive narratives, augmented reality, and AI-assisted story generation. These advances have blurred the line between passive viewer and active participant, the economics of content creation has made it easy for people to share mulit-model narrivies with ease.
TikTok exemplifies the ongoing shift from passive to active engagement between creators and viewers, eliminating the need for traditional infrastructure, tools, and gatekeeping mechanisms.
We become the stories we tell.
Stories, especially science fiction, influence technological innovation. Works like “Star Trek” and “Neuromancer” inspired inventions like flip phones and the internet. Stories reflect our hopes and fears about technology, guiding its development. They inspire us to push boundaries and navigate the ethical implications of our creations.
Science fiction has often predicted technological advancements that later became reality. The Metaverse from Neal Stephenson’s “Snow Crash,” AI assistants like Siri and Alexa inspired by Isaac Asimov’s “I, Robot,” the concept of cyberspace from William Gibson’s “Neuromancer,” and tablet computers and video calls from Arthur C. Clarke’s “2001: A Space Odyssey.”
Authors like Wells, Asimov, Clarke, Dick, Robinson, and Liu have predicted technologies and grappled with their ethical implications. Their stories act as a laboratory for exploring ideas and their consequences, shaping our expectations, innovations, and the future. In a sense these authors are futurists in disguise.
Our narrative preferences mirror our cultural and personal contexts. Stories reflect societal values; for example, climate fiction reflects environmental concerns.
The goal is to create a virtuous cycle: better tools that lead to stronger more impactful stories, which in turn inspire more beneficial real-world innovations, feeding back into even better stories. This approach aligns storytelling tools directly with the aim of influencing the development of positive technology and ideally societal progress. Story telling tools that enable the creative process, not disrupt it.
Make tools that realize and render a stronger future.
Tools focused on promoting stronger, future-shaping narratives.
New storytelling tools, such as generative AI, democratize storytelling by lowering barriers to entry and making it accessible to diverse voices. These tools enable rapid integration of new technologies, generative or otherwise, take advantage of immediate global distribution, and support trans-media storytelling across various platforms and screens. Tools that enhance interactivity, create immersive experiences, and foster collaborative creation and feedback. The opposite of legacy broadcast and film models.
By creating tools that facilitate stronger narratives that everyone can use we can potentially alter the course of our collective future, explore solutions to global challenges, foster empathy, and inspire innovations. Everyone should be able to create using cutting-edge storytelling tools. Maybe we can use power of storytelling and how they impact us to craft a more hopeful, inclusive, and innovative future. The stories we tell and the tools we use are blueprints for the world we wish to build. We all need tools that realize and render a better future.
What happens to music, happens to everyone.
Suno: https://s3.documentcloud.org/documents/24776034/1.pdf
Udio: https://s3.documentcloud.org/documents/24776030/1.pdf
“Accompanying this Complaint and designated as Exhibit C is a thumb drive that contains all the Udio outputs referenced herein and in Exhibit B. In the event Udio seeks to remove this evidence of its infringing conduct from public view, the examples cited herein are preserved on this medium.” (at last this language -ed)
Show the LLM weights, LLM data, open code, or delete the models.
It’s not the the technology is bad, it’s that rights holders did not see this coming, and did not lay the legal frameworks, and industry best practices. The game started without them (again). These are conversations that should have happened years ago, but likely distracted by some metaverse or quantum something, something,.
The declaration from one major label setting boundaries recently as well as this lawsuit, even if was a day late and model short, it’s exactly what I am talking about. Do it.
Prove or remove it should be the remit from all major rights holders in unison and then set the deal terms on a flat playing field, adding on the value of the IP that was stolen and used for to gain market share to the bill. Remember when Google search launched its news area and the news industry did not negotiate a rev split, this, is that again.
Do new deals with rights in tact, auditable deeply embedded rights trackers in latent space for downstream revenues.
Seeing Like A Network- Dark Forests, Dense Networks ( strangeloopcanon.com ) culture is composed of the communication patterns, behaviours, and symbols that are shared amongst a group. We can think of culture as the common interconnected web that underlay the beliefs that we all hold, which constantly changes and evolves as our beliefs spread.
The Future of Streaming (According to Roberts, Malone and Diller) (nyt) Netflix is highly profitable, with operating margins of 28 percent. In the first quarter of 2024, Netflix reported revenue of $9.4 billion, and $2.3 billion in net income. No one else comes close.
Apple Introduces the iStick (spyglass) The EU puked up the carrots, so Apple uses their Intelligence…
OpenAI’s Mira Murati: “some creative jobs maybe will go away, but maybe they shouldn’t have been there in the first place”
AI Doesn’t Kill Jobs? Tell That to Freelancers (wsj) Since the rollout of ChatGPT in November 2022, high-value tasks like IT & Networking have seen pay increases of up to 8%, while low-value tasks such as Admin Support and Writing have experienced significant pay decreases of up to 17% and 18% “When I see something that looks like it was written by AI, I just switch off,” she adds. “The internet has just gotten so much duller.”
I Will Fucking Piledrive You If You Mention AI Again (mataroa) fixed this link “Look at us, resplendent in our pauper’s robes, stitched from corpulent greed and breathless credulity, spending half of the planet’s engineering efforts to add chatbot support to every application under the sun when half of the industry hasn’t worked out how to test database backups regularly.”
Hackers ‘jailbreak’ powerful AI models in global effort to highlight flaws (ft) Machine learning security start-ups raised $213mn across 23 deals in 2023, up from $70mn the previous year
AI is exhausting the power grid. Tech firms are seeking a miracle solution. (wapo) A ChatGPT-powered search on Google, according to the International Energy Agency, consumes almost 10 times the amount of electricity as a traditional search
Calculating Empires: A Genealogy of Technology and Power since 1500 (calculating empires) large-scale interactive visualization exploring how technical and social structures co-evolved over five centuries.
Introducing the next generation of Claude (anthropic) “outperforms its peers on most of the common evaluation benchmarks for AI systems, including undergraduate level expert knowledge”.
AI’s $600B Question (sequoiacap) The AI bubble is reaching a tipping point. Navigating what comes next will be essential.
schelling.ai. open-source, decentralized AI
CDK cyber outage hits US auto dealers for second day in a row (reuters) 15,000 car dealers are now owned and operated by crackers.
We are here: dub a dub a deeeee dub a dub a deeeee dub a dub a deeeee dubbbbla dooo. (Who wants to remix this with Crystal Waters lada di lada do with me ?
Audioseal (meta) Cutting-edge audio manipulation toolkit.
Florence (msft) Microsoft’s advanced conversational AI.
Open-Sora v1.2 (github) Community-driven open-source AI framework.
Video to sound effects (elevenlabs) AI-driven tool for converting video to audio effects.
Generating audio for video (googledeepmind) DeepMind’s AI for creating audio content for videos.
Introducing Gen-3 Alpha (runway) Runway’s latest AI model for creative generation.
Veo: Our most capable generative video model (aitestkitchen) Generating videos.
JACSO (meta) Meta’s smart solution for audio-visual synchronization.
I have this image framed 6×6 massive – orange 60s lucite frame with brown marble border – took off the walls of 51 W 52nd street (CBS/Blackrock) during a refresh. Still have it and its meaning continues to grow.
It’s encouraging to see that the market is moving towards licensing deals for dataset creation, but the current model, where tech companies make one-time payments and never compensate creators again, is fail. Clearly unsustainable by not accounting for the ongoing value generated from the datasets.
Compensation mechanisms cannot be put in place with the current modes because the data was initially trained on unlicensed IP. But it should be clear that no amount of new deals can rectify the fundamental compensation issue, or come close in depth and volume.
Content creators should be actively involved in the dataset creation process, establishing transparent licensing agreements that include provisions for ongoing royalties, rather than one-time payments. By involving creators from the start, allowing them to review and approve the use of their work, and maintaining open communication builds trust through fairness and transparency. Music licensing is a good analog.
Reading list so far this summer. (first half of summer ) see (part 2, end of summer)
All Fours – Miranda July
Co-Intelligence – Ethan Mollusk
The Eye Of The Master
Matteo Pasquinelli
Biography of X Catherine Lacey
Perry Perspective Agents
Daniel Suarez Critical Mass
Charles Duhigg – Supercommunicators
Hari Kunzru Blue Ruin
Kandel -Essays On Art And Science
Tricia Romano – The Freaks Came Out To Wirite
The Twenty Days Of Turin – Giorgio De Maria
Prophet Song – Paul Lynch
Jesmyn Ward Let Us Descend
The Boy, The Mole, The Fox And The Horse – Charlie Mackesy
Candy Darling – Cynthia Carr
Martyr – Kaveh Akbar
The Fraud – Zadie Smith
NYT workplace advice columnist brings it – Goodbye Work Friends….”Still, in my heart of hearts, I always wanted to tell you to quit your job. Negotiate for the salary you deserve. Stand up for yourself. Challenge authority. Tell your rude co-worker to shut up. Report your boss to everyone and anyone who will listen. Consult a lawyer. Did I mention quit your job?” …..
When To Write a Simulator “Any problem involving probabilities over time should humble you. Walk away and quietly go write a simulation.” (sage advice -ed)
SecondPage – Google Searches without media conglomerates (tool) this extension blocks most of the top 1000 English sites from appearing in google search.
Privacy and harm has long been an area of inquiry for me, last week was a bit of a mind blower. Apple put in place secure scaffolding around AI. and understanding (cliaming) your real-time context window with privacy forward AI utility.
OPSEC win for us all. Expect many news cycles of Apple needing to put in a back door. They won’t and never should. Encryption matters and privacy is a fundamental human right.
Amazon-Powered AI Cameras Used to Detect Emotions of Unwitting UK Train Passengers This falls under the worst possible application of ML. You do not get to guess (infer intent) what is in people minds and make that actionable! Full stop. When did the concept of observation implies supervision and therefore control get lost on society. We really are getting the governance we deserve.
While Apple reshaped and recontexualized AI for the marketplace by providing utility around personal privacy, NSA + OPEN AI have decided the exact opposite optics are needed
Treat your data accordingly and expect information governance, negating which data will-and-won’t, be trained and how your data will and will not be used.
After Luma dropped their video offering, the Sora halo holding less power, everyone else is a in constant battle to try and make celloid sense out of latent space. Here is google paper on Generating audio for video. Every day the range of and tools change and improve.
Apple / Huggingface repo – mobile-appropriate vision transformer, CLIP, and image segmentation models
