Google Stitch Just Destroyed Figma's Stock in One Update (Here's Why)
Speakers
Josh
— Host
Ejaaz Ahamadeen
— Co-Host, Limitless Podcast (Bankless)
Summary
- Google's AI design tool 'Stitch' democratizes graphic design, generating full websites from sketches or prompts, which precipitated a 9% single-day drop in Figma's stock, erasing ~$2B in market cap.
- Google also released a new AI coding stack in its AI Studio, directly competing with Anthropic's Claude Code, noted for its browser accessibility and multi-agent capabilities.
- Rivian is criticized for destroying $86,000 of value per vehicle delivered and for a partnership with Uber for 50,000 robotaxis by 2031, despite lacking an autonomy division, a factory for the vehicle, and a realistic timeline.
- OpenAI acquired companies Astral (coding tools) and PromptFu (security tools) to build a comprehensive, enterprise-focused coding AI stack, explicitly targeting Anthropic's perceived moat.
- A disconnect exists between enterprise adoption (73% first-time buyers choosing Anthropic) and prediction market sentiment (94% odds favor OpenAI having the best coding model by March 31st), highlighting narrative power.
- Cursor AI released its own model, Composer 2, but the analysis suggests the accompanying benchmark chart is misleading, with a custom benchmark and inverted axis likely overstating its performance versus cost.
- MiniMax's M2.7 model is reported to have significantly self-improved through autonomous experimentation, handling 30-50% of AI research at its lab, marking a step toward recursive self-improvement.
- Anthropic's global survey of 81,000 people found that concerns about AI reliability outweigh fears of job displacement, with job fears being most pronounced in North America and least in Central Asia.
- The pace of AI competition is accelerating, where single-feature releases or acquisitions can rapidly disrupt established companies and entire market segments.
- Tesla is framed as the superior, vertically integrated competitor in autonomy and EV manufacturing compared to the fragmented partnership approach exemplified by Uber, Rivian, and Nvidia.
Trade Ideas
Rivian lost $3.6B last year on 42,000 deliveries, equating to $86,000 of value destroyed per vehicle. Its new partnership with Uber for 50,000 robotaxis by 2031 lacks a vehicle design, factory, autonomy software, or clear timeline. The company's fundamental economics are unsound, and its ambitious autonomy partnership appears strategically unrealistic given its lack of scale, expertise, and comparable inefficiency versus focused competitors. Rivian's business model is unsustainable without continuous external subsidization, and its new strategic initiative highlights execution risk rather than a viable path to profitability. A new major investor or partner could provide a longer lifeline, or technology breakthroughs could unexpectedly accelerate its timeline.
The speaker states, "My chips are still with Tesla," citing Tesla's vertical integration and execution ability as superior to the partnership model of Uber, Rivian, and Nvidia. In the context of autonomous EV taxis, a single, vertically integrated company with control over its supply chain and software is presented as a more competitive and scalable model than a consortium of separate partners. Tesla is positioned as the favored incumbent in the autonomy/EV space due to its integrated approach and proven scaling capability, which contrasts with the challenged execution of rivals like Rivian. Competitors may successfully execute their partnership model, or Tesla may face its own operational or regulatory setbacks.
Up Next
This Bankless video, published March 20, 2026, features Josh Kale, Ejaaz Ahamadeen discussing RIVN, TSLA. 2 trade ideas extracted by AI with direction and confidence scoring.
Speakers: Josh Kale, Ejaaz Ahamadeen · Tickers: RIVN, TSLA