Anthropic, Palantir, Amazon Collaborate on Defense AI for U.S. Intelligence Agencies,
AMD Joints the party with its own open source AI model and VentureLAB makes a major announcement advancing the semiconductor industry in Canada.
Welcome to Hashtag Trending. I’m your host, Jim Love. Let’s get into it.
Anthropic, Palantir, Amazon Collaborate on Defense AI for U.S. Intelligence Agencies
Anthropic, Palantir, and Amazon Web Services (AWS) have joined forces to offer advanced AI models to U.S. intelligence and defense agencies. The collaboration centers around Anthropic’s Claude models, which will now be integrated into Palantir’s data analytics platform and hosted on AWS’s highly secure Impact Level 6 (IL6) infrastructure.
This partnership will allow agencies to process vast amounts of data quickly, enhance insights, identify patterns, and streamline document reviews. By embedding these AI tools into government workflows, the agencies aim to improve decision-making in critical situations, while preserving human oversight.
This collaboration highlights the growing trend of AI adoption in public sector operations, especially in high-security environments where contractors must meet stringent accreditation requirements. Anthropic, a competitor to OpenAI, benefits from Amazon’s investment, while Palantir’s established position in defense makes it a key player in the AI transformation of government operations. This move aligns with a broader trend, where tech companies like Meta and Scale AI are developing AI solutions for defense, such as the Pentagon’s “Defense Llama.”
AMD Announces OLMo, Its First Open-Source Large Language Model
Following on the heels of NVIDIAs new open source AI, AMD has introduced OLMo, its first fully open-source large language model (LLM), targeting both data centers and smaller organizations. Trained in-house, OLMo offers advanced reasoning and conversational capabilities similar to other LLMs like OpenAI’s GPT-4. The model can run on AMD Instinct MI250 GPUs as well as PCs equipped with Ryzen AI processors and neural processing units (NPUs). AMD also provides AI infrastructure for developers through its AMD Developer Cloud.
OLMo is designed to be flexible, allowing organizations to customize the model by adding their own data during training. The initial OLMo model features 1 billion parameters and was trained using a cluster of Instinct GPUs, processing over 1.3 trillion tokens. The training involved three stages, culminating in alignment with human preferences to ensure that the model reflects human values.
Benchmarks show OLMo performing well in comparison to other open-source models like TinyLlama and MobileLlama, particularly in areas like coding, mathematics, and instruction-following. It also met responsible AI standards, scoring well on tests that measure truthful statements and toxic language.
AMD has made OLMo available for free download, offering a powerful tool for developers and businesses looking to leverage AI without proprietary restrictions.
There’s a link to the download in the show notes.
Download link: https://huggingface.co/amd/AMD-OLMo
ventureLAB and Hudson Valley FastFab Collaborate to Boost North American Semiconductor and Advanced Manufacturing Ecosystem
ventureLAB, a key player in Canada’s hardware technology and enterprise software community, that has helped over 6,800 entrepreneurs raise more than $420 million in capital, has partnered with Hudson Valley FastFab Inc. (HVFF), that provides advanced solutions for the semiconductor and microelectronics industries. The two organizations have signed a Memorandum of Understanding (MOU) to foster innovation in flexible manufacturing, engineering, and workforce development within the semiconductor sector.
This collaboration aims to strengthen the North American semiconductor ecosystem, spanning both Canada and the U.S. Through ventureLAB’s Hardware Catalyst Initiative Lab and HVFF’s expertise, the partnership will support semiconductor startups and scale-ups across the Northeast, offering resources like advisory support, IP and investor readiness, lab infrastructure, and access to corporate networks. The collaboration will also introduce prototype and early-production run capabilities.
A key aspect of the partnership is the expansion of the **HardTech Industrial Cluster™**, an ecosystem dedicated to growing the semiconductor and advanced manufacturing sectors. By bringing together startups, academia, the public sector, and industry partners, the initiative will help fuel the commercialization process for innovative technologies and contribute to the growth of Canada’s knowledge-based economy.
Hugh Chow, CEO of ventureLAB, expressed excitement about the partnership, stating,
And that’s our show for today.
Reach me at editorial@technewsday.ca
I’m your host Jim Love, have a marvelous Monday