Snowflakes CEO Sridhar Ramaswamy interview
Here are the contents of the document translated into English:
This article is a conversation between Ramaswamy and the Chinese magazine “China Science and Technology Innovation”, discussing the development trend of artificial intelligence (AI) and its impact on enterprises.
Ramaswamy first mentioned the application of machine learning and deep learning in the insurance industry, as well as the interest shown by CEOs when automating some processes with these technologies. He believes that if AI’s practicality can be demonstrated, enterprise acceptance will be high.
He then talked about the historical development of technology, such as IBM’s and Microsoft’s deal, DEC’s disappearance, which showed the rapid pace of innovation. Ramaswamy also shared his own experience of leading a team to increase mobile income from 10% to 100%, demonstrating that enterprises can achieve success by actively embracing innovation when faced with new technologies.
Regarding the current investment craze in AI, he said it could be either a good or bad bubble, just like the telecom bubble in the 1990s. Although many companies failed, they ultimately laid the fiber optic cables for Google, Facebook and other companies to develop later.
He also mentioned the advantages of OpenAI and ChatGPT at the consumer end, as well as the possibility of specialized models appearing on the enterprise side, which might not have a single dominant entry point.
Finally, Ramaswamy concluded that in five to seven years, what kind of model development will occur? Will there be many specialized and verticalized models, or will fewer but more general and larger models emerge?
This article, through Ramaswamy’s conversation, shows his profound understanding of AI development trends and enterprise innovation speed, as well as the potential risks and opportunities in current investment craze.
Translation
本文是拉马斯瓦米与《中国科技创新》杂志的一次访谈，讨论了人工智能（AI）的发展趋势及其对企业的影响。
访谈中拉马斯瓦米首先提到了机器学习和深度学习在保险行业的应用，以及通过这些技术自动化部分流程时CEO们表现出的浓厚兴趣。他认为，只要能够展示出AI的实用性，企业对AI的接受度就会很高。
他接着谈到过去科技发展历史，如IBM与微软的交易、DEC的消失等，都体现了快速的创新速度。拉马斯瓦米自己也提到了他带领团队将移动收入占比从10%提升到100%的经历，证明了企业在面对新技术时可以通过积极拥抱创新来取得成功。
对于目前AI领域投资的热潮，他表示，有可能是好泡沫，但也有可能是坏泡沫。就像90年代的电信泡沫一样，虽然很多企业失败了，但是最终为世界铺设了光纤，为后来谷歌、Facebook等公司的发展奠定了基础。
他也提到了OpenAI和ChatGPT等巨头在消费端的优势，以及在企业端可能会出现各种专业化模型，不太可能有一个单一的入口点统治一切。
最后，拉马斯瓦米总结道，未来五到七年后模型的发展将是什么样呢？是充满许多专业化和垂直化的模型还是会出现更少但更通用且非常大的模型呢？
本文通过拉马斯瓦米的访谈，我们可以看到他对AI发展趋势和企业创新速度的深刻理解，同样也看到了当前投资热潮的潜在风险和机遇。
Reference:
https://www.forbes.com/sites/robtoews/2024/12/22/10-ai-predictions-for-2025/## Snowflakes CEO Sridhar Ramaswamy interview
Here are the contents in English:
This is an interview article with Sridhar Ramaswamy, discussing topics such as artificial intelligence (AI), structured and unstructured data, intelligent platforms, and automated insurance underwriting.
The article first introduces several key points mentioned by Ramaswamy during the interview: how robots combine structured and unstructured data to create intelligent platforms, and how CEOs have shown strong interest in automating parts of the insurance underwriting process using these technologies. It then recalls a few significant technological changes in history, such as IBM’s merger with Microsoft, the mobile internet era, and analyzes the experiences learned by these companies during previous transformations and how they are embracing innovation today.
Ramaswamy believes that bubbles will eventually burst, just like various historical bubbles. He references Ben Thompson and Nat Friedman’s discussion on “Is this a good bubble?” and recalls the telecom bubble in the 1990s, as well as companies that burned large amounts of money until Instacart emerged 15 years later. When discussing current AI investments, he says it’s still difficult to determine whether it’s a good or bad bubble.
When discussing model development over the next five to seven years, Ramaswamy analyzes Google’s experience and believes that consumers will tend towards giants like ChatGPT and OpenAI, while enterprise users may see various specialized models emerging, unlikely to have a single dominant entry point.
Finally, the article mentions the subsequent Q&A session and thanks viewers for watching, with a promise of seeing them next time.
Translation
这是斯里达尔·拉马斯瓦米的一篇访谈文章，讨论了人工智能（AI）、结构化和非结构化数据、智能平台以及保险承保的自动化等主题。文章首先介绍了拉马斯瓦米在访谈中提到的几个关键点：机器人如何结合结构化和非结构化数据创建智能平台，以及通过这些技术自动化保险承保的部分流程时CEO们都表现出了浓厚的兴趣；接着，文章回顾了科技发展历史上的几次巨大的变革，如IBM与微软的交易、移动互联网时代等，并分析了这些企业在过去的变革中吸取的经验以及他们如何在如今的AI时代积极拥抱创新。
拉马斯瓦米认为，泡沫终会破裂，就如同历史上的各种泡沫一样。他提到本·汤普森和纳特·弗里德曼关于“这是一个好泡沫吗”的话题，并回顾了90年代的电信泡沫，以及2000年代初那些做杂货配送服务的公司烧了大量资金直到15年后Instacart出现。对于当前的AI投资，他表示现在还很难判断它是好泡沫还是坏泡沫。
在讨论未来五到七年后的模型发展时，拉马斯瓦米结合谷歌的经历分析，认为消费端会倾向于像ChatGPT和OpenAI这样的巨头，而企业端则可能出现各种各样的专业化模型，不太可能有一个单一的入口点统治一切。
最后，文章提到了后面的快问快答环节，并感谢大家的观看，我们下期再见。
Reference:
https://www.youtube.com/watch?v=3-2-w7bPqtk