The Future of Work: How YC Founders Are Enabling Humans with AI

AI-enabled software is changing the way we work. Companies built from scratch in 2023 now integrate new LLM capabilities into their tech stack, widening the possibilities for what young startups can solve for and reimagining ways to create value for their customers.

This is seen most prominently with YC’s Summer 2023 batch, whose demo day this week saw nearly 220 companies presenting. 70% of these companies are building software solutions for other companies (B2B SaaS), and the majority are incorporating some AI features to varying degrees.

With such an influx of AI-enabled startups, how do we distinguish the potential of one from the other? How do we evaluate their impact on the specific departments and professions they’re marketing themselves to?

‍

How AI products can shift work to be more human-centric

New software applications built with large language model (LLM) capabilities aim to handle low-level tasks across various professions; by automating some tasks, workers may refocus their efforts on more human-centric activities such as strategic thinking, empathy, and collaboration, depending on one’s roles and responsibilities.

This is the heart of the human enablement approach to AI, one that creates an entirely new division of labor without devaluing the human worker. This summer, we dove deep into the effects of AI on jobs with our Summer Kapor Fellow, Yasmin Abdulhadi, and we identified three potential ways AI applications can enable humans to work more effectively:

• Automation: AI does the tedious, low-ROI tasks for me
• Augmentation: AI is expanding or extending my own abilities to do my job
• Alignment: AI helps me better coordinate with my team, department, and wider organization to row in the same direction and reduce human coordination issues

Applying the “Three A’s” to top pre-seed and seed companies (YC S23)

Despite being a limited sample size of the overall market, Y Combinator (YC) has attracted some of the highest caliber builders since its launch in 2005. It has helped accelerate the development of household names like Stripe, Airbnb, Dropbox, and GitLab.

Each batch indexes some of the top pre-seed, early-stage companies worldwide that foreshadow new trends in technological solutions. Thus, it is a valuable benchmark for what to expect from the startup market regarding AI enablement.

This batch is particularly unique because it is the first to access the APIs of LLM companies, such as OpenAI’s ChatGPT — granting big-tech software capabilities to early-stage startup companies.

A recent report estimated that 75% of AI’s value-generating use cases come from software engineering and go-to-market functions. This trend is reflected in the B2B SaaS space as well, where many YC founders are building solutions targeting one of the six following roles with high potential for AI integration:

Go-to-Market (GTM): Sales, Marketing, and Customer Operations

Engineering, Product, Design (EPD): Software Engineering, Product, and Design

Below, we highlight how some of YC’s S23 batch is enabling workers through automation, augmentation, or improved alignment and coordination within their organizations.

These are just a few examples of companies that are reimagining the future of work. Many other companies out there are doing innovative work in this space.

Hyperbound | Sales – Prospecting and Outbounding | "Generate high quality, reliable personalized emails at scale that don’t need human review."

Unhaze | Sales – Prospecting and Outbounding | "Unhaze finds and prioritizes the accounts you are most likely to convert so your reps can close deals instead of working through unlikely leads."

HockeyStack | Sales – Prospecting and Outbounding | "HockeyStack improves sales outcomes for B2B companies by helping them understand every interaction their customers have with their marketing, sales reps, CSMs, and products."

Roundtable | Sales – Market Research | "Roundtable makes user and market research faster and cheaper by instantly simulating responses to online surveys."

Raz | Marketing – Customer Intelligence | "Raz provides businesses with LLM agents they can use to automate conversations and build customer relationships."

Subsets | Marketing – Market and Customer Intelligence | "Enabling commercial teams to create AI-driven retention experiments without engineering support.”

Humanlike | Customer Success – Customer Support | "We help companies automate their phone operations using AI voice bots that sound and act human."

Hyperlight | Customer Success – Customer Support | "Our AI agents talk to your customers with intelligence and empathy. Give your customers the support they deserve across voice, chat, and email."

Sweep | Software Engineering – Code Debugging | "Sweep is an AI assistant to automate your software chores like bug fixes and minor feature requests when you spin up a ticket so you can focus on the complex engineering problems."

Metoro | Software Engineering – Code Debugging | "Metoro acts like an experienced engineer. It understands your system and debugs your services one by one, updating its hypothesis as it discovers new information."

Metalware | Software Engineering – Coding | “Metalware is building a copilot for the 1.5 million firmware engineers worldwide. We help hardware companies launch embedded systems 10x faster by automating low-level programming."

Structured Labs | Software Engineering – Maintenance | "Helping developers analyze logs with AI using plain English. Structured lets developers query and analyze log data in English using LLMs. LLMs bring semantic context to logs. For the first time, developers can see the story around previously disjointed masses of log data."

Sudocode | Software Engineering – Coding | "Build software with natural language. Use AI to turn your design docs into code."

Talc | Software Engineering – Planning | "Talc makes it easy for companies to launch reliable AI apps faster. We provide a staging environment that runs through hundreds of realistic scenarios to give developers a clear picture of how well their AI is fulfilling business needs."

Inari | Product Management – Project Management | "Explore product ideas, refine business decisions, and save time drafting docs with a ChatGPT-like assistant connected to your apps and custom to your workflows."

AutoEmber | Design – Prototyping | "Our goal is to make a tool that allows people to create fully-featured web apps without coding."

Tempo Labs | Design – Prototyping | "Tempo empowers anyone to generate and edit high-quality react code using natural language prompts and a visual code editor. Helps teams move 10X faster by eliminating the design→code translation process."

‍

Conclusion

With a new paradigm shift in technology, we are reassessing the division of labor between AI and human workers. As founders build AI-native products, we advocate for a human-centric approach that positions the worker at the heart of the solution. Some guiding questions for this approach can include:

• How can we build products that enable productivity as well as improve quality of life for workers?
• How do we use technology to augment our human capabilities and help us be more effective at our jobs?
• How can a product foster better collaboration within teams, departments, and the broader organization?

Our next post will discuss the ongoing battle of “David vs. Goliath,” and how startups can build defensibility against incumbents as they scale. Subscribe to our newsletter to be the first to see it.