In the generative AI era, how can hardware and software firms innovate?
The unprecedented adoption of ChatGPT has everyone (start-ups, venture firms, and established companies) spending significant resources on assessing how generative AI can be implemented. The fact is, that if they haven’t put resources in place to understand how to improve their processes with generative AI, they will be left behind.
A few ways that both hardware and software companies can innovate is by (i) incorporating generative AI tools in their internal processes and (ii) making enhancements to their offerings through generative AI.
An example of one way in which we have started to see generative AI improve internal processes that is near and dear to Intel is that several Electric Design Automation (EDA) software vendors are adopting generative AI techniques that will help automate and speed up the chip design process. Additionally, we are starting to see AI tools help map the most efficient ways to manufacture within complicated supply chains. Similarly, software engineers can leverage tools like code copilots to help with generating and testing code. 46% of all code written today is generated by GitHub Copilot, making the developers that use it 55% more productive.
Generative AI tools can also help transform functions like marketing, sales, and customer success by writing blog posts, copywriting, image generation and sorting. AI can then ensure that the collateral is targeted at a specific customer type. Intel Capital portfolio company Commonsense Machines has incorporated generative AI techniques to help its users create expansive 3D interactive worlds through natural language text, a feature that would have required days of manual modeling a few years ago. Figure, another Intel Capital portfolio company, has also made innovative steps in AI via its next generation humanoids. Figure 01, Figure’s current robot, is being designed initially for standard yet complex warehouse and logistics facilities to support companies during unprecedented labor shortages. The robot has already achieved a significant milestone by taking its first step.
Additionally, another Intel Capital portfolio company has created an AI function that allows customers to create high-quality content in new languages. This allows them to reach new users in a way that was not available.
There are several other use cases for generative AI that myself and the Intel Capital team anticipate will become common in the coming years including:
- Search: Large Language Models (LLMs) will be used to search within a company’s database to easily surface relevant information and contextualize that information.
- Summarization: LLMs excel at condensing lengthy passages of text into concise summaries that capture essential information such as customer feedback, training materials, and legal terms.
- Content Creation: LLM-based applications are already being employed to create marketing copy and other creative assets. Increasingly, other types of content including requirements specifications, training manuals, test scripts, technical documentation, and audio and visual media could end up being generated.
Companies that are slow to adopt these new technologies will be left behind by those who think creatively about how to implement AI into their workflows. We are at the beginning of the new paradigm and only the paranoid survives.