The Washington Post launches AI chatbot for climate queries

Hello and welcome to this week's newsletter! We're thrilled to bring you the latest and most exciting developments in the world of AI and technology. Let's dive right in:

Try AI as it should be: effortless, powerful, and private

Unlock the real value of AI with local, ownable, and customizable models that you control. Trained on your data. Built for your business. Completely private, and ready to scale at your speed and budget.

• Create: Integrate, customize, 
and deploy AI seamlessly.

• Orchestrate: Quickly connect components, 
assign workloads, and deploy.

• Scale: Leverage existing infrastructure to scale AI applications.

Get Started

The Washington Post launches AI chatbot for climate queries

Why it matters: The Washington Post has introduced an AI-driven chatbot designed to answer user questions about climate, leveraging its extensive archive of climate reporting. This move signals the Post's commitment to integrating AI into its news delivery.

Zoom in: Dubbed "Climate Answers," the chatbot is a collaboration between the Post's product, engineering, and climate editorial teams. It draws on multiple articles from the Post to provide comprehensive responses.

• The AI behind Climate Answers is powered by collaborations with OpenAI and Meta's Llama, utilizing a large-language model tailored to the Post's climate content.

Between the lines: The chatbot is cautious about the information it provides. If it cannot confidently answer a query based on vetted Washington Post reporting, it will refrain from offering a response. This approach ensures that all answers are backed by reliable journalism.

• Users can report any misleading answers via a feedback form, maintaining a layer of accountability and continuous improvement.

Zoom out: The Washington Post is expanding its use of AI beyond the chatbot. Last month, it began experimenting with AI-generated article summaries. Currently available on about 10% of its stories, the goal is to eventually offer summaries for all articles.

• These initiatives predate the arrival of the Post's new CEO, Will Lewis, but his leadership has accelerated the push towards AI integration.

The big picture: The Post is mindful of the potential impact of AI summaries and chatbot answers on user engagement with full articles. Chief Technology Officer Vineet Khosla notes that research shows younger readers often prefer summaries over headlines to decide whether to delve deeper into a story. This trend suggests that AI tools could enhance engagement rather than detract from it.

What to watch: Unlike some other media outlets, the Post has not yet entered into content licensing agreements with AI firms. Khosla emphasized that while they are open to discussions that expand their journalism's reach, any partnership must be fair and beneficial to the Post.

a16z is trying to keep AI alive with Oxygen initiative

Why it matters: Andreessen Horowitz (a16z) is making a bold move in the AI space by securing thousands of Nvidia H100 GPUs, planning to expand to over 20,000. This initiative, called "oxygen," aims to support AI startups by providing essential computational resources.

• Resource acquisition: a16z has acquired thousands of GPUs, with plans to reach 20,000, renting them to portfolio companies.

• Strategic use: Startups receive GPUs at discounted rates or in exchange for equity, ensuring they have the power needed for AI development.

• Notable participants: Luma AI, a multimodal AI startup, benefited early from this initiative.

The context: AI development requires significant computational power, and smaller startups struggle to compete for these resources. a16z’s initiative provides these vital tools, setting it apart from other venture capital firms.

Industry impact: This strategy positions a16z as a crucial partner for AI-driven companies, offering more than just financial support by directly addressing resource constraints.

Future prospects: With the increasing demand for AI, a16z’s substantial investment in GPUs is likely to attract more high-potential startups, reinforcing its leadership in AI investments.

Andreessen Horowitz’s GPU acquisition highlights its commitment to dominating the AI sector by providing critical resources to startups, securing its position in the AI investment landscape.

Ex-Meta scientists unveil massive AI model for protein design

Why it matters: EvolutionaryScale, founded by former Meta scientists, has launched one of the largest AI models for biology, aimed at revolutionizing protein design. This AI model, ESM3, has already created new fluorescent proteins and secured $142 million in funding.

• EvolutionaryScale's ESM3: This protein language model was trained on over 2.7 billion protein sequences and structures.

• Demonstration: ESM3 successfully designed new fluorescent proteins, showcasing its potential.

• Funding: The company raised $142 million to further its applications in drug development, sustainability, and more.

The context: AI's application in biology is expanding rapidly, with EvolutionaryScale joining other innovators using advanced machine-learning models to design proteins. These models can potentially transform medicine by creating proteins tailored to specific functions.

The impact: New Proteins: ESM3's ability to generate novel proteins can lead to breakthroughs in various fields, from medicine to environmental sustainability.

Funding: Significant investment indicates strong confidence in the potential of AI-driven biological research.

Future prospects: EvolutionaryScale aims to use ESM3 for diverse applications, including developing plastic-eating enzymes and protein-based drugs. The model's open-source version allows researchers worldwide to explore its capabilities.

EvolutionaryScale's ESM3 represents a frontier in AI-driven protein design, promising transformative applications across multiple industries. With substantial funding and innovative technology, the company is poised to make significant contributions to biotechnology.