Skip to content
#AI

[2/2] Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI (AI column)

#AI #ML #Trends #Software #Engineering #Future

Continue. survey Mark's speeches will share theses from the second part of his story.

6️⃣ Pricing: usage vs value-based Usage-based – ideal for startups (no upfront fixed costs)transparently But vendors would like to go to value-based - percentage of productivity (For example, the salary of an engineer / doctor). Percentage of upsell/churn reduction in CRM Interestingly, Mark explains that the high prices ($200–300/mes) It's not a bug, it's a feature: it speeds up R&D, reduces customer acquisition waste. AI companies are more creative than SaaS in pricing.

7Incumbents vs Startups: Who Wins? Startups Win in Apps/Domains (Faster, more focused.)but new incumbents (Anthropic, xAI) Other startups that are catching up with1 year. Apps like Cursor aren’t just wrappers – they build their own models. (10–100+), backward integrate.

8A16z Approach: Invest in Conflicting Strategies When a company has fundamentally open questions (big vs small, open vs closed) - It's a problem. (If you guess wrong, you may go bankrupt.). VCs have the advantage of betting on multiple strategies in parallel - "aggressively investing behind every strategy that has plausible chance."

9Society and the workforce: panic in surveys, adoption in practice Polls show panic ("AI will kill the job.")But everyone uses AI. Mark's prediction is that through 5–20 People will say thank you for AI. (as with electricity/ plumbing). So we have to look at adoption. 5 With billions of users online, AI is integrated everywhere from user support to autonomous machines.

For product creators, the whole story can be summarized into these theses.

  1. We're in the beginning. (3 yearly 80) Products are radically evolving.
  2. The AI economy is improving (costs down, revenue up) Don’t be afraid of costs if there is a real demand.
  3. Small models catch up - you can build without frontier infra.
  4. Open-source is a game changer – China has proven it can be done with less resources.
  5. Regulation - risk No1 (watch out for states/EU).
  6. Pricing is creative – don’t be limited to the SaaS model.
  7. Competition is tough, there is no permanent lead - speed is critical.

#AI #ML #Trends #Software #Engineering #Future