I agree what you’ve listed makes sense as a product portfolio.
But AI Studio is getting vibe coding tools. AI Studio also has a API that competes with Vertex. They have IDE plugins for existing IDEs to expose Chat, Agents, etc. They also have Gemini CLI for when those don’t work. There is also Firebase Studio, a browser based IDE for vibe coding. Jules, a browser based code agent orchestration tool. Opal, a node-based tool to build AI… things? Stich, a tool to build UIs. Colab with AI for a different type of coding. Notebook LM for AI research (many features now available in Gemini App). AI Overviews and AI mode in search, which now feature a generic chat interface.
Thats just new stuff, and not including all the existing products (Gmail, Home) that have Gemini added.
This is the benefit of a big company vs startups. They can build out a product for every type of user and every user journey, at once.
In "real world" you don't use OpenAI or Anthropic API directly—you are forced to use AWS, GCP, or Azure. Each of these has its own service for running LLMs, which is conceptually the same as using OpenAI or Anthropic API directly, but with much worse DX. For AWS it's called Bedrock, for GCP—Vertex, and for Azure it's AI Foundry I believe. They also may offer complementary features like prompt management, evals, etc, but from what I've seen so far it's all crap.
Jules is the first and only one to add a full API, which I've found very beneficial. It lets you integrate agentic coding features into web apps quite nicely. (In theory you could always hack your own thing together with Claude Code or Codex to achieve a similar effect but a cloud agent with an API saves a lot of effort.)
Startups aren't supposed to be "venture scale" when they launch. (I don't know what "venture scale" means, but it sounds big.) They're supposed to make something that at least a few people want, and then iterate.
just keep in mind you weren't in the pitch room. I'm old enough now to realize that not everything we see in public is the full story.
It may be all a pipe dream and not pan out, but I bet they pitched a path to more broader optionality. That's all you really need, momentum and optionality.
Saw it on Twitter and was interested. But from the video and demos I immediately did not understand why Notebooks and Notes are two tabs? In my mind, a Note is IN a Notebook, not some separate adjacent item...
This is, as far as I understand, self healing ONLY if the name of a CSS class changes. Not for anything else. That seems like a very very very very narrow definition of "self healing": there are 9999 other subtle or not so subtle things that can change per session or per update version of a page.
If you run this against let's say a typical e-commerce page where the navigation and all screen elements are super dynamic — user specific data, language etc. — this problems becomes even harder.
My running hypothesis on this is that AI is a sentient screenreader and the last thing you should be worrying about is CSS class names, IDs, data-testid attributes, DOM traversal, and all of these things that are essentially querying the 'internal state' of a page. Classes, IDs, data attributes, etc. aren't a public API and semantic elements, ARIA attributes, etc. are.
So, focus on WCAG compliance, following the spec as faithfully as you can. The style or presentation of something may change as part of a simple A/B test but the underlying purpose or behaviour would remain the same.
I feel like this could work if the selectors are chosen carefully to capture semantic meaning, rather than basing off of something arbitrary like a class name. The agent must have some understanding of the document to be able to perform those actions in the first place.
If it can find an ellipse tool, it's likely based off some combination of accessible role, accessible name, and inner text (perhaps the icon if it's multi-modal.) So in theory, couldn't it capture that criteria in a JS snippet and replay it?
That's exactly what is it doing. The workflows are pretty much js snippets in itself you can see in the "code" tab (in the plugin when you select a saved workflow).
Everyone thinks of typical e-commerce pages when its comes "browser agent doing something", but our real use cases are far from shopping for the user. But your point still stands valid. The idea is that maybe there are websites where generating stable selectors/hierarchy maps wouldn't solve, but 80% (from 80-20) of websites are not like that (including a lot of internal dashboard/interfaces) (there will also be issues for websites with proper i18n implementations if the selectors are aria label based)
Self healing css selectors is also only 1 part of the story. The other part is the cohesive interface for the agent itself to use these selectors.
Why do you have an accountant? Or a lawyer? It's the same thing. Corporations don't have all skills in house for a ton of things.
I was an IT consultant. A big energy company wanted to go to the AWS cloud. Their folks were too busy and had no experience. We (my consultancy company) already had the knowledge.
Consultants don't only give advice. In many cases, they also do the work. But advice is also a "product". If your in-house team does not have the knowledge or time, you hire a consultancy firm.
From my understanding, Browserbase mostly provides remote browsers for their users. We also provide remote browsers, but with a lot more infrastructure on top (single/multi web agents, caching/reliability support, an agentic IDE/editor, etc).
not really sure how that differentiates since those things you mentioned are ancillary to main value. Also - browser base is insanely cheaper, but looking at the prices this doesn't look like a real company mainly just a way to have users in free tier (with toy level limits)
reply