I find it hard to believe there aren't worse ones out there. But off the top of my head I actually can't think of any...

Quite incredible that you can say a sentence like that, while talking about something within Google(/Alphabet) that's actually an important area for them (not just some random little side project), and yet it isn't hyperbolic, it's sincere!

Everyone there overseeing it (from direct management all the way up to C-suite) should truly feel embarrassed.

Claude used to allow editing prompts which created branches, but imo it was a much crappier version of branching versus what I could do in AI studio. Alas, Anthropic have even removed that functionality for some reason (if someone knows how to branch in Claude please lmk).

Sometimes I just want to block off all web search and sample from LLM’s knowledge only. I guess I can tell Claude to not search the web, but I prefer doing it deterministically like AI studio.

I also like that I can set a different system prompt per session, and all my system prompts get autosaved for future use.

I find it quite surprising that with all this AI, my history of sessions is still such a mess despite my efforts to keep it organized (in both AI Studio and Claude). At least Claude finally added folders, but they are still buggy (for sessions older than a few months, for example).

https://x.com/thsottiaux/status/2083596911060324570?a=iif1W7...

Their backend looks very confused too. I consume their models through OpenRouter and requests are processed randomly by AI Studio or Vertex. API key management had several entry points for a while too, with the option to query the same model with a GCP, Firebase and/or AI Studio API key.

When I signed a few devs at work up for Gemini for Code I had to go into Google Cloud, jump through several different screens, assign special permissions to each dev to make it work and even when it did work...it was insanely confusing compared to every other option out there.

It's to the point where I'm better off asking Grok Bot set it up for me rather than ever deal with figuring it out again.

Is this true anymore? I feel like a lot of the best talent have better opportunities elsewhere where they won't get bogged down by entrenched bureaucracy. Not to mention the Google interview process and physical locations filter out a lot of talented people from even wanting to apply.

No longer true. Not since CFO Ruth Porat IBMified the compensation policy in the name of efficiency. Btw this is exactly you can't get away with underpaying tech talent - something new will come along that threatens to eat your lunch and if your team doesn't have the best people, you're finished. And you can't reverse reputation erosion overnight - part of the attractiveness of taking a high-paying job is that everyone knows you're being paid high.

This. There are so many hoops to jump, so many links to click, so many pages to read. I just don't know why people would even bother with it.

My guess is that some GCP's VP won the argument that Gemini must go through AI Studio. These VPs don't use their product, so they don't give a shit about UX. As for their engineers and product managers? Well, that beats me.

The effort to fix it vs. the risk seems weirdly asymmetrical. Why do all the hard things right and then trip up at the last hurdle?

You also have to hire for taste.

Do they still have 5 messaging apps or did they finally trim it down.

unfortunately, the so-called "best talent" is being hired in management, not engineering or design or UX. :-(

This could be used as a preface for so many things.

It's been a baffling question for over a decade (or more)

Right. They lack no resources. Therefore, what is happening must be what they want to happen.

So, not that easy for me (I should note im not in the USA)

Maybe it's a b2b vs b2c thing.

So, no, it is not astroturfing.

"How do I even pay for it?" is particularly baffling to me considering it's literally just included in their normal range of Google One subscriptions that bundle Drive/etc which have been around for years.

You pick a plan and pay for it, and sign in to whichever app you're using like Gemini chat (ChatGPT/Claude Chat) or Antigravity (Codex/Claude Code).

https://one.google.com/about/

AI Studio is the developer console (OpenAI Platform/Claude Platform). Vertex AI is the Google Cloud enterprise platform (using GPT or Claude models via AWS Bedrock/Azure AI Foundry).

Pretty much every product is a 1:1 equivalent with the same offering from OpenAI/Anthropic + Azure/AWS. Google is both a hyperscaler and an AI lab, so the venn diagram includes both the direct OpenAI/Anthropic products and what they also serve for enterprises via Azure/AWS.

I get that Google does bad stuff and isn't popular on HN which is understandable, but the specific criticisms really don't make sense. The entire AI developer/consumer space is messy and confusing but that's an industry wide complaint not specific to Google unless they should just not compete on certain verticals.

Antigravity come with Google's models and anthropic - so Gemini 4 Argon, Gemini 3.8 Flash, Claude Opus 5.5 Claude Sonnet 5.5 and GPT-OSS 120B.

Agy is better than Claude code in some ways and vice versa.

Google models are great.

Ironically google music is so terrible I've uninstalled it and use Metrolist exclusively.

I think google models do a decent job helping me put together a blog, for example:

https://hypertekst.net

(Content is typical 90s homepage stuff, not great. But model voice isn't as grating as Claude - and while it has a "look" I think it's s bit better).

Would want assurances they won’t permaban a Gmail for a false positive for something on their LLM offering. Like the risk with Google Photos (thinking about the dad sending a doctor a photo of his son’s rash).

Have used them both for work with google workspace accts, and personally, with antigravity in linux.

Luna/Terra/Sol/Astra was a bit pretentious but at least the sizing makes sense. (I imagined the meeting... "Ok guys, we're getting mogged on naming... what's more epic than a Myth?" -"The SUN!")

Anthropic makes sense with the literary "scale" naming, mostly.

Is a Carbon bigger than an Argon though? Maybe it has to do with the atomic number?

I have been a pixel owner since the OG pixel. I am furious with how Google has killed assistant which was fast and accurate and could operate on device if I did not have internet and replaced it with Gemini in a way that forced it down my throat and removed existing functionality for something worse that I can only use for the majority of useful use cases if I allow them to train on my data and use it for ads and measurement.

Those are product decisions that they did not have to go with.

It has completely soured me on Gemini given how much they sold my current pixel 10 Pro as having the latest in on device AI capabilities.

It is to the point where I have invested substantial tokens in building alternative solutions that I sideload onto my phone via codex to replace basic things like voice interaction with home assistant.

They have dropped the ball in so many ways and not all of them are related to the actual model quality so much as product and business decisions.

This narrative seems astroturfed particularly hard on this website. Like Google's AI isnt allowed to be praised on HN.

https://vercel.com/ai-gateway/models/gemini-omni-1.2-flash

If you're using any sort of router middle-man new models are usually easy to try.

DuckDuckGo’s equivalent feature is unfortunately too prone to hallucination, especially whatever cheap model they are using for AI summaries. Once you click in it switches to GPT Luna, which will sometimes hilariously correct the previous response.

It’s like the difference between me looking at a picture, describing it to you, and then you trying to form that picture in your mind. You might get close, but it will never be the picture I’m looking at. If I then ask you questions about the picture, you can’t answer any questions about any details that I didn’t describe. I believe that is exactly the information and understanding loss that an Omni model attempts to address, which doesn’t have anything to do with output.

Some models do also offer “Omni” output, but that is more like a feature of the model than what Omni models are by design I think.

Omni 1.2 is showing 18 credit cost in Flow for a six second video.

vs.

10 credits costs for a 6 second video using Omni 1.1.

I'm not sure it's in Google's interest to talk about how much cheaper the model is, while also increasing the cost in Flow.

(but I'm happy for people using the API)

I think they are doing a terrible job with their product packaging. Like others, I also gave up with Google models long ago due to difficulty in accessing and low limits.

My gut is that this is reflective of their fragmented internal product organization around AI.

Go with the Flow, don’t ride the Wave.

Google is just beyond me I have no idea what they're doing, and no interest in figuring it out. I use a free gemini flash model in one of my personal apps, but that's it

For B-Roll it's different. I can somewhat see the why behind that. It's just to give your brain something to look at so that you don't have a voice-only podcast.

But movies OTOH are these creative endeavors where the shots _are_ the content instead of just augmenting the content.

This won't stop the industry from still trying to use this tech like that, of course.