This Week in Startups
This Week in Startups

Demoing Google Suite's AI plus legal docs, detection, and more! | E1835

This Week in Startups is brought to you by… Lemon.io. Get access to Lemon Hire, a platform with more than 80,000 pre-vetted engineers that you can interview within 48 hours. Get $2000 off your first hire at http://lemon.io/hire today! Codecademy. Build the future you want to see with Codecademy. Cod

Featured Speakers

Jason Calacanis Host

Topics Discussed

Episode Summary

Executive Summary: The hosts survey startup opportunities in the Middle East, then demo AI tools for legal documents, spreadsheet analysis, AI-content detection, video summaries, and Google Workspace writing. They see strong regional startup potential and regard AI-assisted drafting as already useful, while noting uneven reliability: Google Docs performed well, but spreadsheet generation and some data analysis failed. They also discuss model access, AI’s role in everyday browsing, and consumer technology.

Main Topics: Middle East startup and venture capital opportunity: Sunny Madra describes a fast-changing startup ecosystem in Saudi Arabia, the UAE, and Egypt, including local services tailored to regional needs and sophisticated family offices. He predicts the region could become a major venture capital center as technology adoption and customer markets grow. AI tools for startup legal work: LegalNow uses AI and document templates to generate and revise common startup agreements, such as intern contracts, NDAs, equity plans, and terms of use. The hosts see its value in lower cost and customization, while emphasizing that higher-stakes legal work still requires care. AI analysis of spreadsheets and content: Rows offers natural-language questions and automated insights for spreadsheet data, but its demonstrated analysis sometimes returned incorrect or incomplete answers. Originality.ai identified a Google AI-written performance review as 100% AI-generated; the hosts questioned how the detector works. AI writing in Google Workspace: Google Docs' 'Help me write' feature drafted and revised a performance review and a response requesting a raise. The hosts praised its ability to overcome writer's block and speed up routine writing, contrasting its strong performance with weaker Google Sheets generation. AI summaries and future interfaces: Summarize.tech creates skimmable, section-by-section summaries of long YouTube videos and can reuse summaries already generated by others. The hosts suggest summaries could become a native YouTube feature and could offer follow-up questions or deeper context. AI models, devices, and everyday use: The discussion considers whether OpenAI's consumer product uses newer or different models than its API, and whether AI assistants could evaluate information while users browse. The hosts also discuss Apple's hardware and voice-based ChatGPT use for conversational learning.

Key Arguments: Regional startups can grow by adapting global business models to local conditions; the hosts cite services such as Breadfast and Careem as examples of locally relevant offerings. AI can make common startup legal documents cheaper and easier to customize than traditional law-firm work or static templates, but users should remain alert to legal risk and errors. AI spreadsheet tools have a promising natural-language interface, yet incorrect answers and weak generated outputs show that reliability remains a major limitation. AI-assisted drafting can materially reduce time spent on routine writing when users provide the substance and intent; the Google Docs demo produced a usable draft and revisions quickly. AI detection may be useful for editors and readers, but the demonstrated score alone does not explain the detection method or establish how dependable it is. AI features may be more useful when embedded in existing workflows—such as Docs, YouTube, or a browser—rather than requiring users to copy and paste between separate services. A strong consumer AI product may combine multiple models and API-level orchestration, so comparisons with a standalone model may not be direct; the speakers present this as an informed possibility, not a confirmed account of OpenAI's systems.

Data Points: Breadfast regional bread business: About $100 million - Madra describes the startup as doing roughly this amount in bread sales in the region. Saudi Arabia population: 35 million - Mentioned as the scale of the local customer market. Egypt population: 100 million - Mentioned as a large potential market for products and services. Startups housed at The Garage in Riyadh: 150 - Madra describes the coworking space as housing around 150 startups. Forecast for the Middle East's venture capital position: Number two within 10 years - Madra's prediction for the region's ranking among venture capital investment regions. Originality.ai detection score: 100% AI-written - The tool assigned this score to a performance review generated by Google's AI. Video summarized by Summarize.tech: 3 hours 45 minutes - A Lex Fridman podcast featuring Jared Kushner; the service presented a section-by-section summary. Google Sheets conference-planning prompt: 3 days and 150 executives - The hosts tested whether the tool could create a conference budget including food, lodging, and entertainment. Lemon Hire candidate pool and interview availability: 80,000+ engineers; interviews within 48 hours - Figures stated in the episode's sponsor message.

Pivotal Quotes: "The startups here and the founders are kind of like I would say New York in the 90s" — Sunny Madra: He compares the region's current startup ecosystem with an earlier stage of New York's development. "The crazy thing about this is if you actually know your points to put them in full sentences, this does a wonderful job." — Jason Calacanis: He summarizes the strength of Google Docs' AI writing tool when the user supplies the ideas. "This is the 'don't worry, be crappy' mode that we're in with AI. You got to build stuff and you got to throw it out in the world" — Jason Calacanis: He frames the rapid release of imperfect AI features as part of the current development cycle, after Google Sheets failed several prompts.

Implications: AI is already useful for routine drafting and information synthesis, but users should verify outputs—especially legal advice, analysis, and generated budgets. Products that fit existing workflows may win adoption. Meanwhile, regional market growth and localization could create substantial startup and investment opportunities.

🔓 Sign Up for Unlimited Episode Search

About This Week in Startups

Jason Calacanis covers startups, tech, markets, media, and all the hottest topics in business and technology. He also interviews the world’s greatest founders, operators, investors, and innovators.

View all episodes from This Week in Startups