Hacker Newsnew | past | comments | ask | show | jobs | submit | ukuina's commentslogin

To the sibling comments: Apple holds <10% of the worldwide PC market: https://www.gartner.com/en/newsroom/press-releases/2025-01-1...

Congrats on the move to Google!

Please allow me to rant to someone who can actually do something about this.

Vertex AI has been a nightmare to simply sign up, link a credit card, and start using Claude Sonnet (now available on Vertex AI).

The sheer number of steps required for this (failed) user journey is dizzying:

* AI Studio, get API key

* AI Studio, link payment method: Auto-creates GCP property, which is nice

* Punts to GCP to actually create the payment method and link to GCP property

* Try to use API key in Claude Code; need to find model name

* Look around to find actual model name, discover it is only deployed on some regions, thankfully, the property was created on the correct region

* Specify the new endpoint and API key, Claude Code throws API permissions errors

* Search around Vertex and find two different places where the model must be provisioned for the account

* Need to fill out a form to get approval to use Claude models on GCP

* Try Claude Code again, fails with API quota errors

* Check Vertex to find out the default quota for Sonnet 4.5 is 0 TPM (why is this a reasonable default?)

* Apply for quota increase to 10k tokens/minute (seemingly requires manual review)

* Get rejection email with no reasoning

* Apply for quota increase to 1 token/minute

* Get rejection email with no reasoning

* Give up

Then I went to Anthropic's own site, here's what that user journey looks like:

* console.anthropic.com, get API key

* Link credit card

* Launch Claude Code, specify API key

* Success

I don't think this is even a preferential thing with Claude Code, since the API key is working happily in OpenCode as well.


You went further with GCP than I did. I was asked repeatedly by support to contact some kind of a Google sales team.

I get the feeling GCP is not good for individuals like I. My friends who work with enterprise cloud have very high opinion about their tech stack.


> I get the feeling GCP is not good for individuals like I.

Google isn't good for individuals at all. Unless you've got a few million followers or get lucky on HN, support is literally non-existent. Anyone that builds a business on Google is nuts.


I'd like to state the AWS, in contrast, has been great to me as an individual. The two times that I needed to speak to a human, I had one on the phone resolving my issue. And both issues were due to me making mistakes - on my small personal account.

I propose a new benchmark for Agentic AI...Be able to sign up for a Google Service...

Yes, it’s extremely complicated. I gave up on fire base for one project because I could not figure out how to get the right permissions set up and my support request resulted in someone copying and pasting a snippet from the instructions that I obviously had not understood in the first place.

It’s also extremely cumbersome to sign up for Google AI. The other day I tried to get deep seek working via Google’s hosted offering and gave up after about an hour. The form just would not complete without error and there was not a useful message to work with.

It would seem that in today’s modern world of AI assistance, Google could set up one that would help users do the simplest things. Why not just let the agent direct the user to the correct forms and let the user press submit?


Oh man, I've been playing with GCP's vertex AI endpoints, and this is so representative of my experience. It's actually bananas how difficult it is, even compared to other GCP endpoints

Then you actually use it! I dare someone to try and get Gemini live vertex app working.

I wonder if OpenOMF has the same limits.

It's a keyboard thing and less of a software thing.

This is so neat looking. Is there an equivalent for MacOS?


Not exactly afaik, but I've recently been going to System Settings > Accessibility > Display, and turning on:

    Increase contrast
    Reduce transparency
    Differentiate without color
    Show toolbar button shapes
https://imgur.com/a/DqfN07k

I like the retro and simple vibe compared to the new Liquid Glass controls.


Ah! Thank you! Even on Sequoia this is a massive improvement!


Great, glad to help. FYI there are similar settings for iOS and I do same on my phone.


At 0:52 in their demo video, there is a grammatical inconsistency in the agent's text output. The annotations in the video are therefore suspected to be created by humans after the fact. Is Google up to their old marketing/hyping tricks again?

> SIMA 2 Reasoning:

> The user wants me to go to the ‘tomato house’. Based on the description ‘ripe tomato’, I identify the red house down the street.


I can't speak to the content of the actual game being played, but it wouldn't surprise me if there was an in-game text prompt:

> "The house that looks like a ripe tomato!"

that was transformed into a "user prompt" in a more instructional format

> "Go to the tomato house"

And both were used in the agent output. At least the Y-axes on the graphs look more reasonable than some other recent benchmarks.


The scene just before you describe has the user write "ripe tomato" in the description - you can see it in the video. The summary elides it, but the "ripe tomato" instruction is also clearly part of the context.


Very much this.

You are better off asking it a write a script to invoke itself N times across the task list.


Same. I think there’s an untapped market (feature really) here, which if isn’t solved by GPT-next will start to reveal itself as a problem more and more.

LLMs are really bad at being comprehensive, in general, and from one inference to the next their comprehensive-ness varies wildly. Because LLMs are surprising the hell out of everyone with their abilities, less attention is paid to this; they can do a thing well, and for now that’s good enough. As we scale usage, I expect this gap will become more obvious and problematic (unless solved in the model, like everything else).

A solution I’ve been toying with is something like a reasoning step, which could probably be done with mostly classical NLP, that identifies constraints up front and guides the inference to meet them. Like a structured output but at a session level.

I am currently doing what you suggest though, I have the agent create a script which invokes … itself … until the constraints are met, but that obviously requires that I am engaged there; I think it could be done autonomously, with at least much better consistency (at the end of the day even that guiding hand is inference based and therefore subject to the same challenges).


The "loaded question" approach works for getting MUCH better pro/con lists, too, in general, across all LLMs.


> if LLMs "knew" when they're out of their depth, they could be much more useful.

I used to think this, but no longer sure.

Large-scale tasks just grind to a halt with more modern LLMs because of this perception of impassable complexity.

And it's not that they need extensive planning, the LLM knows what needs to be done (it'll even tell you!), it's just more work than will fit within a "session" (arbitrary) and so it would rather refuse than get started.

So you're now looking at TODOs, and hierarchical plans, and all this unnecessary pre-work even when the task scales horizontally very well (if it just jumped into it).


Is it really worth the time, though?

You are better off allowing your overworked neural pathways some much-needed rest.


That question was asked 8 years ago. Coincidence? I think not!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: