More Popcorn Please!

Good morning.

I don’t know what’s going on at OpenAI but we need some more popcorn.

Another three very high profile people announced their resignation including CTO Mira Murati. At the same time Altman is going around telling people superintelligence is only about 1000 days away. WHAT’S GOING ON ??

I’m getting some slight Elizabeth Holmes vibes. We demand answers ;)

Anyway , back to our original programming.

A few days ago I met someone I went to high school with.
We hadn’t seen each other for 30 years.

The first thought I had was, Oh my god, this is an old man. Not the nicest thought, I admit, but in my defense, I was a bit stunned by his appearance.

In my mind, this person was still a fresh-faced 17-year-old.

He might have thought the exact same thing by looking at me. The strange thing is that as we look into the mirror every day, we don’t really see ourselves age. But you get the occasional reminder if you come across an old photo or meet someone from your past.

The same goes for technological revolutions. You don’t know that it’s happening if you’re in it except for subtle (little crow’s feet appear near your eyes) or not-so-subtle hints (you cannot drink 4 beers anymore without paying the price).

Let’s take the internet as an example.

I grew up in a world without internet. I was playing around with bulletin board systems in the early 90s on one of these modem lines. And I remember when arriving at university in 1995 I had to cycle across town to get an email address. The internet were a few hundred pages , that’s it. (Perhaps I AM an old man 😉 )

And now 30 years later - it has so thoroughly changed everything we do.

I also remember the skepticism against the internet. Unimaginable now, but here’s a good example:

And then you have the visionaries - like David Bowie (rest his soul) in 1999. What he says in the clip below seems obvious now but is nothing but a stunning piece of insight coming from that era.

So why am I showing you this? Because the same thing is happening. Only a lot faster.

You will no doubt have met the people who call AI an alien life form, and you will have encountered the people who think it’s all bullcrap.

I think about this a lot from a professional standpoint.

Suppose you’re a copywriter in 2024.

You might be in denial and think that computers can never do what YOU do. Oh no, you go around and you call yourself a professional writer. An artist. But with words. You can put feeling into your writing, you can show empathy and soul, and computers are nothing but empty soulless boxes.

But the reality is that large language models are doing a good enough job and the economic reality is that companies simply will take the cheapest option so for a fraction of your hourly rate your client can get a ChatGPT subscription and will replace you. No questions asked.

Your client will also not tell you. Your phone will simply stop ringing and your email box will stay empty.

You might be in denial and keep going.

But between competition from platforms like Upwork and Fiverr ( where everybody is also using ChatGPT) but for 5$ per hour as opposed to your 100$ the reality is that you’re out of a job.

Your pride will make sure that you won’t see it coming and don’t know when you are already defeated.

And it’s not just copywriters, it’s programmers, graphics designers, video editors,... You’re all up for disruption. No , scrap that. We are ALL up for disruption.

So what to do ? You have only one option: embrace AI and lean into it. Contact your client and perhaps teach them and their employees on how to use AI.

Pivot!

If the world changes around you, you have to adapt and evolve. It’s how we got here as a species.

Every time i talk about AI I meet this person. The person that says he tried ChatGPT and is not convinced by the output. Head in the sand.

But know that these models get better and better.

Keep learning and stay informed. Knowledge is Power.

But you’re here and that’s exactly what you’re doing. Keeping up to date!

If you want to throw this lifeline to other people in your life, use the referral code underneath the email and get some Amazon dollars.

Welcome to the Blacklynx Brief !

[End of Rant]

By the way, there is now a referral program for this newsletter. For every 3 people that you get to sign up, we will gift you a 15€ Amazon gift card.

Scroll to the bottom of the newsletter for details!

(Open)AI News

• Mira Murati, OpenAI's former CTO, has announced her departure after six and a half years, amidst speculation that the company is shifting away from its non-profit structure. Murati plans to focus on personal exploration while ensuring a smooth transition, aligning with reports that OpenAI is restructuring into a for-profit corporation, with CEO Sam Altman gaining equity.

• OpenAI is expanding its Advanced Voice Mode (AVM) to all ChatGPT Plus and Teams users, offering new voices and improved features for more natural AI interactions. The updated AVM now includes personalized conversation recall and better understanding of accents, but will not be available in certain regions like the EU and UK. This upgrade aims to make daily AI interactions feel more human and seamless.

• At Meta's Connect 2024 conference, the company unveiled several AI advancements, including the new Orion AR glasses, Llama 3.2 AI models, and AI-powered updates to Instagram and messaging platforms. The Orion AR glasses feature voice control and hand tracking, while Llama 3.2 introduces advanced image and text understanding with smaller versions for mobile devices. Meta also rolled out AI tools for Reels, automatic video dubbing, and a new Voice Mode for chatting with Meta AI across its platforms.

• Google has launched new versions of its Gemini AI models, featuring faster performance, better math abilities, and lower costs for developers. The updates include models with a 20% boost in math tasks, pricing cuts over 50%, and faster output with reduced latency. These improvements make AI development more affordable and efficient, helping developers create smarter applications.

• James Cameron, the renowned director, has joined the board of Stability AI, the company behind the Stable Diffusion AI model. Cameron views generative AI as a key player in the future of visual media, aligning with Stability AI's focus on creative technology. His involvement highlights Hollywood's evolving relationship with AI, as filmmakers explore its potential amid ongoing industry debates.

• At T-Mobile's Capital Markets Day, OpenAI CEO Sam Altman highlighted the new o1 model's advanced reasoning skills, marking a major step forward in AI problem-solving. Altman hinted that o1 is still in an early stage, comparable to GPT-2, but that AI development is accelerating, with agent-based systems potentially arriving soon. This suggests a rapid leap from o1 to more advanced AI models, possibly within months. See the video below

• In his personal blog, Sam Altman predicts that superintelligent AI could emerge in a few thousand days, revolutionizing human progress and capabilities. He foresees AI tools solving complex problems, personalized virtual teams of experts, and breakthroughs in healthcare and education. Altman believes this "Intelligence Age" could drive unprecedented prosperity but will require vast computing power and energy to be widely accessible.

• Microsoft has signed a 20-year deal with Constellation Energy to reopen the Three Mile Island nuclear power plant, aiming to power its data centers and AI operations with clean, carbon-free energy. The agreement highlights the growing energy demands of AI and tech companies’ push for sustainable solutions. Set to reopen in 2028 as the Crane Clean Energy Center, this move aligns with broader industry trends, including Bill Gates' investment in nuclear energy.

• Former Apple design chief Jony Ive has confirmed his collaboration with OpenAI CEO Sam Altman on developing an AI-powered hardware device. The device aims to offer a more seamless, less disruptive computing experience than the iPhone, with Ive’s design firm LoveFrom leading the project alongside former Apple executives. With plans to raise up to $1 billion in funding, the project seeks to push the boundaries of AI hardware by handling complex user interactions.

Quickfire News

• Warner Bros. Discovery adopted Google Cloud's AI for caption generation to reduce production time and costs in unscripted programming.

• Suno AI introduced a cropping feature for AI-generated songs, allowing Pro and Premier users to adjust the start and end of their creations.

• Apple announced plans to release Siri's major AI-powered updates gradually, with the biggest updates expected in iOS 18.3, likely arriving in January 2025.

• Intel launched its Xeon 6 processors and Gaudi 3 AI accelerators, claiming doubled performance for AI workloads and better price-performance compared to Nvidia's H100.

• OpenAI expanded API access for o1 models, adding tier 4 users at 100 requests per minute and increasing tier 5 users to 1000 requests per minute.

• Duolingo added AI-powered Adventures mini-games and a Video Call feature to improve language learning through practical and immersive experiences.

• LinkedIn sneakily introduced AI model training using user data, with users needing to opt out via account settings and a separate form to stop future data use.

• MIT researchers developed an AI model called Crystalyze to determine structures of powdered crystalline materials using X-ray diffraction data.

• The UN’s AI advisory board made seven global governance recommendations, including creating a scientific panel and establishing a global AI fund.

• Amazon launched Amelia, an AI assistant for third-party sellers in the US, helping with account issues and sales data, currently in beta.

• TikTok is developing its own AI GPUs, planning mass production of two chips for AI training and inference by 2026 to reduce reliance on Nvidia.

• Salesforce will offer free access to its premium AI courses and certifications via the Trailhead platform through 2025.

• Disney researchers developed an AI system that teaches robots complex movements using random motion data, improving their versatility for various industries.

• Google introduced new evaluations for AI models handling long-context reasoning, with their 1.5 Pro model outperforming others by 12.5% on advanced tasks.

• Middle Eastern sovereign wealth funds have increased their AI startup investments fivefold over the past year.

• Alibaba’s Qwen 2.5 7B coder model reportedly matches GPT-4’s coding abilities at a lower cost, offering open-source alternatives.

• Pudu Robotics unveiled PUDU D7, a semi-humanoid robot with 30 degrees of freedom and an 8-hour battery life, designed for service and industrial use.

• Researchers developed AnySkin, a tactile sensor for robots, allowing them to apply learned manipulation techniques across various sensor models.

• Researchers developed an AI capable of solving Google’s reCAPTCHAv2 image-based tests with 100% accuracy, challenging the effectiveness of human verification.

• Vice President Kamala Harris voiced support for AI development while ensuring consumer protections during a speech at a New York City fundraiser.

• LinkedIn suspended AI training using UK users' data after privacy concerns raised by the Information Commissioner’s Office, pausing the practice for further review.

• Andy Serkis teased a new project featuring AI-generated characters that move from 2D to augmented reality.

• OpenAI launched Academy, offering $1 million in API credits, training, and support to AI developers in low- and middle-income countries.

• OpenAI introduced the Multilingual MMLU dataset, translating the benchmark into 14 languages to improve AI models' knowledge across diverse cultures.

• Meta announced it will not immediately join the EU's voluntary AI Pact, focusing instead on complying with the upcoming AI Act.

• OpenAI is working on an upgraded version of its Sora AI video generation model, targeting improved quality and longer video clips.

• The Federal Trade Commission (FTC) took enforcement actions against several companies for deceptive or unfair AI usage.

• Nvidia analysts projected the production of around 450,000 Blackwell AI GPUs in Q4 2024, potentially generating over $10 billion in revenue despite early challenges.

• Nebius Group plans to invest over $1 billion in AI infrastructure across Europe by mid-2025, focusing on GPU clusters and data centers.

• OpenAI CEO Sam Altman confirmed the early completion of the Advanced Voice Mode rollout for ChatGPT, except in regions requiring external review.

Closing Thoughts

That’s it for us this week.

If you find any value from this newsletter. Please pay it forward and receive 5€ for every person you get to sign up in the form of an Amazon gift card. What’s keeping you ?