Sep 18, 2025

Investors Discuss The Language AI Thesis at SlatorCon Two seasoned venture capital investors, Shesh Amathnadu and Pramod Gosavi, joined the Investor Panel at SlatorCon Silicon Valley in early September. The pair shared their view on the language AI thesis and areas of opportunity within language AI with a 200-strong audience. Amathnadu is Senior Investment Director at SK Telecom Ventures, the corporate venture arm of South Korea’s largest telecom operator SK, whose investments include Anthropic, Perplexity, and 12 Labs. Gosavi is Senior Principal at Blumberg Capital, a US-based B2B generalist investor, which makes investments from Pre-Seed to Series B from its two funds. With solid engineering backgrounds and more than a decade of investment experience, both Amathnadu and Gosavi have become experts in AI investing. Amathnadu discussed the current AI cycle and described the launch of ChatGPT in 2022 as an inflection point, which created a “clear differentiation of traditional machine learning [ML] versus generative AI.” In summary of the ChatGPT era to date, he noted that 2023 saw investors ask “what are the applications that are emerging and how can you actually commercialize some of these technologies into the real market?” Whereas, 2024 was “when the whole industry realized that the cost is significant.” As he pointed out, “the big question is where is the ROI for some of these gen AI applications?” Now, in 2025, with the cost of AI decreasing, Amathnadu said, “the way I see it, [the focus is on] more vertical AI, that is very industry-specific and domain-specific.” While AI has offered a lot of promise to a fresh wave of startups, one potential risk is attempting to play in an overly crowded market. A case in point, according to Gosavi: “If you’re just doing translation or transcription, that’s going to commoditize very quickly. Whereas, I think what we’re looking for is, clearly, a use case out of it.” To illustrate his point, he offered the example of product marketing. “Now you can [not only] translate. You can translate and transcribe that video into different languages and you can send it to your customers in different countries. I think that’s a very personalized experience. We are looking for more unique end-to-end products and real use cases.” Gosavi also gave examples of how language AI could help to solve real-world problems in the area of logistics (such as routing for truck drivers or phone support for tenants reporting home repairs). Amathnadu echoed Gosavi’s sentiment around the benefits of personalization and urged tech builders to remain focused on creating real value for end-users. Two scenarios he outlined were the pursuit of low latency in voice applications and enabling a family watching Netflix to experience a multimodal and immersive viewing experience with real-time translation of video, text, and image. The pair also highlighted some potential investment hotspots, citing voice, voice understanding, and video understanding, particularly when coupled with the understanding and transfer of emotion, as exciting areas. “To me, the more interesting part is how can we make progress and get productivity, as a community and as a world?” Gosavi said. It was a fitting remark to end a conference that focused on the evolution of the language industry community amid the progress of AI. Tags