🎉 AI Hardware, RoboTaxis, & Hollywood's Demise By Stock Video

Secret Altman and Jony Ive Hardware Collab, Tesla Robotaxis Coming Soon, Google Paywall, Apple's New AI Model, Will Shutterstock Disrupt Hollywood?

Welcome to the first weekly edition of AImpulse by Jes Bickhart, a five point summary of the most significant advancements in the quickly-evolving world of Artificial Intelligence.

Here’s the pulse on this week’s top stories:

What’s happening: It's reported that Sam Altman, CEO of OpenAI, and Jony Ive, the former lead designer at Apple, are currently engaged in discussions to secure up to $1 billion in funding from prominent investors including Emerson Collective and Thrive Capital for their new company. This venture is shrouded in secrecy but is known to focus on developing a personal device powered by AI.

1. The proposed device will not look like a phone, but will likely integrate with OpenAI’s conversational AI tech.

2. Altman was previously in discussions with Japanese investment giant SoftBank on funding, with news of the hardware ambitions leaking in late 2023.

3. Altman is also an investor in Humane AI, which will ship the first batch of its personal ‘AI Pin’ devices this month.

Why it matters: The collaboration between Altman and Ive marries exceptional design with cutting-edge AI technology, positioning it as a potential underdog in the competition against established hardware behemoths for on-device AI. Yet, this partnership also amplifies existing concerns regarding the increasingly indistinct boundaries between OpenAI and Altman’s expanding network of ventures and interests.

What’s happening: Elon Musk just announced that Tesla will reveal its long-awaited fully autonomous Robotaxi at an event on Aug. 8th, which comes as rumors swirl that the company is abandoning plans for an affordable line of EVs.

1. On Friday, Reuters reported that Tesla's entry-level EV will no longer be built.

   Musk quickly called the Reuters report ‘a lie’, later posting that the Robotaxi will be unveiled on Aug. 8.

2. Musk previously said the company’s Robotaxis would be operational by 2020 and would feature no pedals or steering wheel.

3. Tesla just celebrated hitting the milestone of 1 billion miles driven by its fully self-driving (FSD) vehicles.

Why it matters: Tesla has a grand vision for the future of transportation — and AI advances coming from Musk’s companies suggest the tech could be closer than most believe. But given legal battles and inclination for bold timelines, we’ll see when the autonomous rubber actually meets the road.

What’s happening: Google is reportedly considering charging for new ‘premium’ AI-powered search features, marking the first time the company would put any of its core search engine products behind a paywall.

1. Google is developing tech to deploy AI-enhanced search as part of its premium subscription services, which already includes Gemini access.

2. The traditional search engine would remain free, but certain AI-powered search enhancements would be limited to subscribers.

3. AI is costly to run compared to Google's current search model, potentially threatening the company's $175B search advertising cash cow.

Why it matters: Google's consideration of a paywall underscores the conflict between its current ad-supported revenue strategy and the transformative impact of AI. Introducing a paywall might counterbalance rising expenses, but it also threatens to lose market share to rivals such as Microsoft, who have quickly adopted AI into their search services at no extra charge.

What’s happening: In a new research paper, Apple researchers introduced ReALM, a new AI system that can understand on-screen tasks, conversational context, and background processes.

1. ReALM uses a new approach of converting screen info to text — allowing it to bypass bulky image recognition parameters for more efficient on-device AI.

2. The model takes into account both what's on the user's screen and what tasks are active.

3. According to the paper, Apple's larger ReALM models substantially outperformed GPT-4, despite having fewer parameters.

Example use case: If scrolling through a website and you want to call a business, you could tell Siri to “call the business“, and Siri would be able to “see“ the phone number on the website and call it directly.

Why it matters: ReALM is a big step forward in making voice assistants more context-aware. By understanding on-screen info and additional context, the next Siri update could provide a more seamless and hands-free user experience.

What’s happening: A new report from Reuters revealed new details on how tech giants like Google, Meta, OpenAI, and Apple are racing to secure vast quantities of online data to feed their AI models.

1. Reuters reports that Meta, Google, Amazon, and Apple all reached a deal with Shutterstock in 2022 following ChatGPT’s debut.

2. The agreement included hundreds of millions of images, videos, and music files for AI training, with deals ranging between $25-50M.

3. Prices for training data range from cents per image to hundreds of dollars per hour of video.

4. Companies are also paying for access to private content archives, including Photobucket's 13B photos and videos and other old internet platforms.

Why it matters: The significance lies in the shift from merely scraping the web to tech giants now investing heavily in premium content for AI model training, spurred by numerous lawsuits and the deluge of AI-generated content online. This investment is likely to heavily influence the development of foundational models for longer-form video LLMs as well.

Will the commonality of training data catalyze a disruption in Hollywood, or will the generative works lose their appeal due to a lack of novelty? I’m torn but was encouraged by the short films released by OpenAI’s Sora model so far. Let me know what you think!

---