It’s an attempt by a researcher from DeepMind named Arthur Mensch, as well as former Meta AI scientists Timothée Lacroix and Guillaume Lample to challenge the hegemony of OpenAI. They have decided to build an open source AI solution that will “Make AI useful.” It plans to release its first models for text-based generative AI in 2024.
While this is the biggest development that has happened outside Silicon Valley, the harsh regulatory environment of the European Union (EU) can be a major hurdle on the path of the AI company. For instance, unlike other parts of the world, OpenAI hasn't been able to make inroads in the EU because of the strict data rules of GDPR and EU AI ACT and has the meagre user base of 3.98% of the total users globally.
Meanwhile, the development will ruffle a few others who are against the idea of AI development in open source space. OpenAI’s Sam Altman has raised concerns about it as the open source models can be used by bad actors and harm humanity. As opposed to Altman’s view, the developer community supports the open source AI cause and have used Meta’s LLaMa and Stability AI’s to build fine-tuned models like Alpaca, Vicuna and others.
While the ecosystem celebrates the inflow of funds in the open source AI space, it’s too early to predict what this company aims to achieve, but it has definitely laid a path for others in the AI space.
Read the full story here.
Apple’s Infiltrates Brain via Vision Pro
Apple has taken a significant step in the realm of brain-computer interfaces (BCIs), a concept that has long been confined to science fiction. Known for their groundbreaking innovations, Apple has filed a patent called 'eye-gaze based biofeedback' for their Vision Pro headset. This patent explores the non-invasive detection of human thought by combining neurotechnological solutions with machine learning algorithms.
By tracking a user's attentive state while consuming specific content, Apple aims to enhance XR experiences by utilizing the gathered biofeedback data. This development showcases Apple's commitment to pushing the boundaries of technology and revolutionizing the way we interact with devices.
Read the full story here.
That’s how Purplle Hires Data Scientists
Purplle distinguishes itself in the industry by leveraging AI and analytics to enhance the customer experience and achieve operational excellence. Their 11-member team, comprising seven data scientists and four data analysts, plays a pivotal role in driving these advancements. Purplle seeks individuals with strong analytical capabilities, programming skills, and a deep understanding of statistics and data visualization.
The interview process consists of two rounds for junior data scientists and three rounds for senior data scientists, assessing technical proficiency, problem-solving abilities, domain knowledge, project experience, leadership skills, and cultural fit. Candidates are evaluated based on key result areas, including project impact, code quality, error resolution, innovation, and performance optimization.
Read the full story here.
In AI Chip Battle, AMD Rises to Occasion
AMD has unveiled the MI300X, a new GPU designed specifically for AI tasks, in an effort to compete with NVIDIA's dominance in the AI accelerator market. With up to 192GB of memory, this chip is well-suited for handling large models like LLMs. This move comes as NVIDIA's market capitalization hits $1 trillion, making it crucial for AMD to establish its presence in the growing AI market.
AMD showcased the chip's capabilities by successfully running Falcon 40B entirely in memory. Additionally, AMD updated its RocM platform, offering a competitive alternative to NVIDIA's CUDA language. The MI300X's ample memory bandwidth also makes it an appealing choice for smaller companies with lighter AI workloads.
Комментариев нет:
Отправить комментарий
Примечание. Отправлять комментарии могут только участники этого блога.