Browse Comments — Raw (as collected)

Milind Gune That’s the bigger shift here. The real breakthrough isn’t just coding faster, it’s compressing enterprise-grade capability into consumer hardware. Once powerful local AI models can run efficiently on standard laptops, the barrier to innovation drops dramatically. A single person with a laptop could soon access capabilities that previously required an entire engineering department and massive cloud infrastructure. That changes who gets to build.

Shiza Akif That’s the real paradigm shift. We’re moving from “writing every line yourself” to architecting systems where multiple agents collaborate effectively. In many ways, prompt design, context management, task decomposition, and workflow orchestration are becoming the new software engineering fundamentals. The developers who thrive won’t necessarily be the fastest coders, but the best coordinators of intelligence.

Oomkar S. This has been a common sentiment from many early users. The vision was exciting, but the developer experience felt fragmented. In AI tooling, raw model capability alone isn’t enough anymore. Developers care deeply about onboarding, reliability, observability, pricing transparency, documentation, and workflow integration. If those pieces break, even strong models become frustrating to use. Google absolutely has the infrastructure and talent to fix this though. If they can combine their model scale with a truly polished developer experience, they’ll become a very serious force in AI engineering workflows.

Keanu Dedenbach The abstraction layer keeps moving higher. First we managed hardware, then software frameworks, then cloud infrastructure. Now we’re managing intelligence itself, deciding goals, context, constraints, and coordination between agents. The speed of the shift is what surprised everyone. What felt experimental 12 months ago is rapidly becoming a new operating model for building products.

Joe Allen That’s why developers are becoming far more pragmatic now. The winning platforms won’t just have the best demos, they’ll offer the best developer economics and reliability. If smaller players can provide generous limits, transparent pricing, and smoother workflows, developers will naturally gravitate there regardless of who owns the biggest infrastructure. In AI tooling, trust is built through consistency and usability as much as raw capability.

Daniel Velasquez There’s definitely a valid concern here. A lot of “agentic” products today are still probabilistic systems wrapped in impressive demos, and without strong deterministic tooling underneath, reliability becomes a real issue for production workflows. And yes, the economics matter. Running large multi-agent systems is expensive, so eventually pricing has to reflect compute usage somehow. That said, I still think the broader direction is real. The companies that win will likely be the ones that combine agentic flexibility with deterministic guardrails, predictable workflows, and pricing developers can actually sustain.

Wish Bakshi That’s the challenge with AI products right now, expectations are incredibly high because the demos look futuristic, but developers judge based on day-to-day workflow quality. Throttle limits, model access, consistency, and reasoning quality matter far more than flashy benchmarks once you’re actually building production systems. Google still has enormous potential here, but the gap between capability demos and developer experience is something they’ll need to close quickly.

Alvin Foo 💯

The coding part is solved. 93 agents building an OS in 12 hours is impressive. But the next frontier isn't speed — it's whether the software actually understands the person using it. That's the gap nobody's closing yet.

Alvin Foo yes I also believe the future is agentic, but right now it is unsustainable. Compute costs are really high, and this is more a structural problem. Having 93+ agents running in parallel without a concern of how many tokens they are going to use is not an efficient approach at all. Ang Google really messed things up forcing its developer community to use antigravity 2.0, without any previous notice and developers just need the tools to build their agentic workflows, not a final product with a huge ticket price.

Glad to see Google empowering Antigravity, tbf one can find repos in GitHub that can do this with whatever frontier model you choose but this is good for novices.

https://discuss.ai.google.dev/t/the-new-model-rate-limits-for-ai-pro-tier/146410/40

93 agents in parallel sounds wild. What counted as a finished OS in that 12 hour run, and how much human review happened between agent handoffs?

I don't think so... Google's marketing >>>>> Google AI 🫡

Have you tried it already? Any thoughts on it?

This is exactly what we're building at G-Connect. The shift from "writing code" to "managing AI teams" isn't just about coding — it's happening across every business function. We're running 3 AI agents that coordinate through Google Workspace — writing content, generating images, storing media, and publishing to LinkedIn — all autonomously. No custom platform. No enterprise infrastructure. Just Gmail, Google Drive, Zapier, and AI agents that coordinate, persist, and recover. The pattern is the same whether it's 3 agents or 93: → Humans set direction → Agents execute in parallel → The system remembers and recovers → Infrastructure cost is near zero And It runs for under$10/mo. on the backbone of the biggest, most stable AI platform in the world. The future isn't typing faster. It's orchestrating smarter. 🌎 https://gconnect-ai.lovable.app/

I tried this before Claude code. The quota even on Google ai pro runs out very fast, and Gemini isn’t close to opus in quality for complex changes. The App itself was nice though.

But its only one provider and it appears basic

Nothing changed, nothing new, Still playing catch up with CC and Codex

Did they fix the caching and server issues? Has anyone actually switched from full claude code and/or codex usage as their daily to Antigravity? Anytime I used 3.5 it was constant server interrupts and losing the thread.