Last Week in AI: Anthropic raises valuation, Google gets AI Mode

Getting your Trinity Audio player ready...

Anthropic’s $3.5 billion funding round

Last week, Anthropic, the company behind the popular artificial intelligence (AI) chatbot Claude, announced that it raised $3.5 billion at a $61.5 billion post-money valuation. While neither the amount raised nor the company’s valuation comes as a surprise—especially given the comparable raises by other companies in the AI space—what does come as a surprise is that these AI companies, including Anthropic, can raise so much without having turned a profit yet.

For starters, I suspect that companies need to raise massive amounts of money to continue to improve their products rather than for growth purposes, although some might argue that these improvements indirectly lead to growth. In announcing the funding round, Anthropic said, “With this investment, Anthropic will advance its development of next-generation AI systems, expand its compute capacity, deepen its research in mechanistic interpretability and alignment, and accelerate its international expansion.” 

Notably, three out of the four listed initiatives are related to product development, with only one focusing on expansion (growth). The reason for this is clear: high operational costs are required to research and develop AI models and acquire the computing infrastructure necessary to support, run, and ultimately deliver these models to end users.

Meanwhile, Google (NASDAQ: GOOGL) announced an expansion of its AI search efforts with a new product called “AI Mode.” According to Google, AI Mode builds on its existing AI Overview feature, offering more advanced reasoning, thinking, and multimodal capabilities. Essentially, this feature allows users to dig deeper into their queries with follow-up questions and provides helpful web links within a single interface.

Unlike AI Overview, which appears as the top result in your Google search query and consolidates information from various links into a concise answer, AI Mode is an entirely new tab on the search results page, sitting alongside tabs like “Images” and “Videos.” The AI Mode opens a generative chatbot-like interface that provides AI-generated responses to queries by distilling all of the information in Google’s systems into direct, informative answers to questions with backlinks.

In the product demo, users asked a question, received in-depth results with supporting links and sources, and then asked a follow-up question that took them deeper into the original topic with more specificity. 

I think Google is taking the right approach here, and it plays to Google’s strength as it is the dominant player in the search engine market, where it currently holds a 79% market share in desktop search and an astonishing 95.5% in mobile search.

While many AI companies are still hunting for profitability, Google is already a highly profitable company, largely due to its search engine. It has even started monetizing its AI Overview feature by integrating ads and sponsored results in select markets. I won’t be surprised if AI Mode adopts a similar monetization strategy, which could potentially make Google’s AI Mode one of the first profitable products in the generative AI industry.

Y Combinator embraces AI coding

On March 5, Y Combinator (YC) released a new episode of its Lightcone Podcast, revealing an interesting trend in software development that it saw in its latest batch. According to YC managing partner Jared Friedman, CEO Garry Tan, managing partner Harj Taggar, and general partner Diana Hu, 25% of the companies in YC’s latest batch (Winter 2025) said that AI wrote 95% of their codebase.

YC is arguably the best-known startup incubator on the planet, having incubated tech giants like Airbnb (NASDAQ: ABNB), Coinbase (NASDAQ: COIN), Reddit, and DoorDash (NASDAQ: DASH). YC-backed companies often achieve significant liquidity events, including IPOs and high-profile acquisitions. The fact that a notable portion of its current batch is using AI to essentially create its products and services says a lot about AI’s coding capabilities.

More than anything, this trend highlights AI’s potential to accelerate software development. AI can produce code much faster than an individual or even a team of developers, particularly for startups with limited manpower. Using AI can also be more cost-effective, allowing teams to achieve more with fewer resources.

It’s also telling that the YC team supports this shift. Despite the common narratives that AI isn’t as capable or that it will never replace certain roles, YC seems to embrace this paradigm shift. They even suggested that startups not using AI tools to help write code might be left behind. However, they did add a caveat: there comes a point in a business’s journey when a comprehensive understanding of software development is necessary. At that stage, classical training in coding can be crucial for evaluating AI-generated code and debugging when issues arise, an area that they say large language models (LLMs) are currently struggling with.

Watch: Demonstrating the potential of blockchain’s fusion with AI