⏺ An AI tool analyzes cancer patients’ facial expressions and outperforms doctors in predicting their prognosis.

Scientists have developed FaceAge, an AI capable of scanning a cancer patient’s face and more accurately predicting short-term survival compared to human doctors. Trained on nearly 59,000 healthy faces and tested on 6,196 cancer patients, the algorithm identifies “biological age,” which is often older than a patient’s actual age. The older the face appeared, the poorer the prognosis.

When doctors relied solely on photos, their predictions for 6-month survival were accurate 61% of the time. However, when FaceAge was introduced, the accuracy increased to 80%. While concerns about bias and confounders remain, it is undeniable that AI has the ability to discern mortality from a person’s face.

@aipost 🪙 | Our X 🥇

Meta AI dropped Meta Perception Language Model, an open & reproducible vision-language AI for challenging visual tasks

It can watch videos and extract details like what a person is doing in the content and how they are doing it

@aipost 🪙 | Our X 🥇

Anthropic launched web search in the API

The move allows developers to build apps where Claude can search the internet for up-to-date info and provide answers with citations

@aipost 🪙 | Our X 🥇

⏺ The CEO of Uber says not enough of his employees know how to use AI: 'Absolute necessity' within a year

Uber CEO Dara Khosrowshahi stated that many of his employees lack the skills to use AI effectively, emphasizing the need for training programs within the company. He called for educational institutions to also teach AI skills, viewing them as essential tools for the future workforce.

Khosrowshahi highlighted that learning to use AI for coding will be crucial at Uber within a year. He noted a trend among tech leaders mandating AI usage in their companies, reflecting a broader industry shift towards integrating AI into everyday work practices.

@aipost 🪙 | Our X 🥇

⏺ Jensen Huang says 100% Nvidia Engineers Will Have AI Agents

Jensen Huang, CEO of Nvidia, predicts that 100% of the company's engineers will utilize AI agents, marking a significant shift in engineering practices. He emphasized that AI tools will enhance productivity by helping with tasks like debugging and optimizing code. Huang described this transition as a redefinition of engineering work itself, where engineers will collaborate with AI agents rather than solely relying on traditional methods.

He acknowledged challenges related to energy and compute power, which could limit AI's widespread adoption. Huang's vision suggests that industries will increasingly integrate AI into their workflows, creating a competitive advantage for those who adapt early.

@aipost 🪙 | Our X 🥇

⏺ Zuckerberg’s Grand Vision: Most of Your Friends Will Be AI

Mark Zuckerberg emphasized the transformative potential of AI in shaping the digital future during a recent event. He highlighted that AI could enhance communication, improve user experiences, and create new opportunities across various industries.

Zuckerberg also discussed Meta's commitment to integrating AI into its platforms, aiming to foster innovation and provide users with more personalized interactions. He acknowledged the challenges of AI implementation but remained optimistic about its ability to drive significant advancements in technology and society.

@aipost 🪙 | Our X 🥇

Sam Altman says AI may be the biggest technological shift in human history, and even he doesn't know where it’s going.

“These tools are capable of things we can't quite wrap our heads around.” Let's approach the singularity with humility and caution, he says. What could go wrong?

@aipost 🪙 | Our X 🥇

Senator: Who’s winning the AI race? and how close is china to us?

Sam Altman: It’s our belief that OpenAI has the best models in the world… it’s very hard to say how ahead we are from china but i’d say.. not a huge amount of time

@aipost 🪙 | Our X 🥇

Wild. This ChatGPT o3 prompt = $20k growth consultant in your pocket 🤯

@aipost 🪙 | Our X 🥇

⏺ AI is now guessing your future diseases in the UK

A new AI called Foresight is meticulously analyzing the health data of 57 million individuals residing in England with the ambitious goal of predicting potential health issues before they even manifest. Essentially, it functions as a sophisticated version of ChatGPT, but instead of assisting with homework excuses, it focuses on predicting heart attacks.

If you’ve ever visited a hospital, you’re likely familiar with the fact that your personal information is often collected. Rest assured, your name is not included in this data. Scientists believe that Foresight has the potential to revolutionize healthcare by preventing strokes and hospital visits before they occur. While this concept may seem unsettling, it holds great promise for improving public health outcomes.

@aipost 🪙 | Our X 🥇

🇺🇸 Google’s plan to fuel AI? Go nuclear or go home

Google has partnered with a startup called Elementl Power to construct three nuclear power plants to power its data centers that heavily rely on AI. Each plant will eventually generate enough energy to power a small city, and yes, Google intends to acquire the electricity generated by these plants.

Although the specific type of reactor has not been chosen yet, the focus is on the fact that AI requires substantial power, and Google is not waiting for the grid to catch up. Instead, it is proactively constructing its own future energy source.

@aipost 🪙 | Our X 🥇

🇺🇸 AI brings murder victim back to court to forgive his killer

Chris Pelkey, a victim of a road rage incident in Arizona, made a remarkable comeback three years later. Thanks to artificial intelligence, he appeared in court, his voice recreated using old videos and voice clips. This digital version of Chris engaged in a direct conversation with the man who had killed him, instead of resorting to anger.

Instead, he delivered a heartfelt message of forgiveness, expressing regret for their encounter and suggesting that they might have been friends in a different life. Pelkey emphasized his belief in forgiveness and the existence of a forgiving God, a sentiment that deeply moved the judge. The killer was sentenced to 10.5 years, while Chris’s brother described the experience of seeing the AI version as a profound sense of healing.

@aipost 🪙 | Our X 🥇

Sam Altman got COOKED

@aipost 🪙 | Our X 🥇

Microsoft dropped Pages, a new Copilot feature that allows users to collaborate with the AI assistant on answers

It works pretty much like ChatGPT Canvas, but doesn't seem to have coding capabilities

@aipost 🪙 | Our X 🥇

⏺ Microsoft and OpenAI may be renegotiating their partnership

Microsoft and OpenAI are reportedly in negotiations regarding their partnership, with discussions focused on Microsoft’s equity stake in OpenAI's planned for-profit entity. OpenAI is restructuring its corporate framework but will retain control under a nonprofit board. Microsoft, which has invested $13 billion in OpenAI, may give up some equity for access to future technology developed post-2030.

The negotiations are complicated by the growing competition between the two companies as OpenAI's enterprise business expands and it pursues its ambitious Stargate infrastructure project.

@aipost 🪙 | Our X 🥇