If that was the business, yes. The first rule of understanding the business is understanding the business model. Not the product. If the business model is a subscription, then you look at the moat. If the moat is not there - you are not that much different from Microsoft Office (which arguably has a larger moat).
Announced the model two weeks ago to a ton of hype, model was unavailable for anyone to try. On twitter, Google folks went on a great run taking shots at Muse ("Muse who?"). The hype died down bc nobody could touch the models and meta/muse came out with a similarly capable model in the meanwhile.
At the same time, OpenAI had a disastrous dev day that led a lot of users to jump ship to Claude.
Couple weeks later, Gemini-4-argon is still unavailable on anything less than the ultra tier, which means there's no natural funnel to get ppl to pay for the more expensive ultra plan other than "trust us the model is good".
If Google had argon broadly available earlier this month they would have picked up a ton of new users.
As someone internal at Google I don't understand why there's so much criticism of Google models... but right now I'm using the next model beyond Argon, and Flash 3.8. In my personal Gemini app I only have access to 3.5.
Models move so fast now and Google is very cautious about releasing them.
My complaint, and the complaint of many is not the models - they can be good to very good depending on use case.
What sucks, hard, is googles ability to deploy direct to devs and customers at scale when the deployment is not a longstanding Google project.
Case in point: I’m an ultra subscriber and see in my app right now an ad to try Gemini 4; meanwhile the only models I have available are 3.1 pro,3.6 flash, and 3.8 thinking.
Clicking the try gemini 4 link takes me to a demo prompt that loads 3.6 flash. Srsly
I won’t even start telling you what getting api key based access together was like a year ago.
I think google makes great models, my favorite models have been google models. My critisisms are with the infrastructure around the models (or lack of them),things like billing, access, changes to terms and limits, pricing etc. All of these have shown either inconsistency, lack of care, sometimes confusing offerings, and things that felt like rug pulls.
Please google, just make it easy for me to give you money and get what I paid for.
I agree that reply was unnecessarily rude, but... idk man. I'm a huge cheerleader for Google [1], use 3.1 Pro via AI studio, I'm on a Pro plan... and I can't access this model? Not even a "4 chats per hour" type thing?? I mean, I have video generation access for some reason (no good for me), but Google's policies are baffling to say the least.
Ofc I know this isn't your fault in person, but I really wish Google would get its shit together.
...As a long time tech enthusiast I should probably stop being so delusional. Heh (cries internally)
1: lol I know who cheers for a trillion dollar company but I'd rather not cheer for OpenAI
There's nothing unnecessarily rude about my comment, they are literally unable to get their act together on product.
The top level comment blamed cheerleading but let's apply critical thinking for a second as product people and look at the funnel:
1. Potential users sees new release
2. Potential user takes time to install and set up Google harness (you already lost 90% of people here, the funnel better close quick!)
3. Potential user cannot use new model.
Then what? You hope they developed good will towards your model through this experience?
You stuck a two week wait in your funnel? Is that the definition of not having your act together or what?
We're on HN, how many of us have heard from Google staff this kind of stuff is a result of them growing so big that AI Cloud has to compete with Vertex Cloud, and AI cloud exists because Vertex Cloud needed 2 years to get OAI/Ant style API keys working and were losing dev mindshare in the meantime?
Because Google models are crap. They are on par with local shit models. Last time I tried they were unable to even hold a longer conversation, because they would enter loop and reiterate last answer regardless of the prompt etc.
Then basic mistakes (inferring wrong information from the given text, hallucinations...). Gemini is okay for some absolute basics and the rest of the work goes to actual frontier models.
How long ago did you try it? Gemini models seem extremely capable to me. I've been using them in ag for over a year and to me the results are great. switching to opus in ag doesn't produce any better results for me
All the power users are always jumping around by now and many hold the smallest subscription with all the models or at least two.
Narrating this as if it’s the game of thrones is a fun exercise, but it’s just becoming commonplace now that models jump ahead of each other on a weekly to Monthly basis.
> and many hold the smallest subscription with all the models or at least two
This was part of the issue, as stated in their comment. We couldn't jump around to try it for the reason that is wasn't made available to, as you say, "all the power users".
I'm still a codex man myself but lots of people feel like opus 5.5 just gets it for some projects in a way that astra doesn't. Especially given the fact that astra and sol 6.1 are often at capacity late at night during US hours, when lots of people want to run longer unsupervised projects while they sleep. There's a general sense I've felt reaching some comment threads that openai failed to react with something better than opus 5.5 and are rushing to release small tweaks to make people feel like they're still getting something new.
Sol is just hilariously slow. Even Luna is slow. It is unusable compared to Claude. I was pretty unhappy with Opus 5 and really gave OpenAI a shot for personal projects, but their quotas are just too small and their capacity issues in the last few weeks have just been untenable. No good. Opus 5.5 is just a very good model, very efficient with usage, gets things done at a reasonable speed. Sonnet 5.5 is also a very nice worker improvement. It is probably on par with Opus 4.6-4.8.
The sense I get from your comment and the GP comment is that OpenAI’s pricing changes alienated individual users for the sake of the enterprise market. I live in the enterprise market, and was quite pleased with the changes. We don’t really use Sol 6.1, just Astra.
$500/month for 40+ hours a week of Astra usage, with ultrafast mode available when you’re near the weekly reset has made our jobs so much easier. Our product has some sciencey bits as well, ands Astra kills it as a rubber duck in this area.
As a consumer whose 20x account got downgraded, I am a bit sad. I can’t justify $500 a month on personal AI spend. As an engineering manager, I could not be happier with OpenAI’s services. Also, Codex App Server is amazing.
I'm a bit confused by this comment. I thought Enterprise customers didn't have access to fixed price plans like the $500/month personal tier and instead paid per token? If that's available to businesses, Astra seems like a no-brainer. But at least the regular Business plans only show me the $25 and $125/month options and the latter is a very recent addition.
Hey, account created 39 days ago. Did you really have to tag that on after all the other stuff? Could you make it any more obvious?
That said this thread is hilarious because it's just an open ai sales manual debating with an anthropic sales manual. It's funny, but it's not, because the truth is that you people are just shitting all over the commons. Someone needs to kick you out but that's not going to happen here.
This product has made my life substantially easier. I have no affiliation with OpenAI.
Claude refuses to help you build your own harness off a subscription, and OpenAI actively encourages it with Codex App Server. So you’re just in an AI IDE customized to your job and specific needs ALL THE TIME. You know that good feeling you got when your CI pipeline was finally working, back when that involved speed reading documentation and guessing, or staying at work until 10PM? You just get that all the time because life is constantly improving. And your tools are beautiful and riddled with Easter eggs you put in them. There’s so much more time to test and talk to stakeholders because you are able to distribute the thing you are actually good at -solving people’s difficult problems- while following best practices. You used to have to weigh those two every day and it was terrible. And everything was always late. You had to cut so many corners you were delivering circles.
Now life is fun and I’m over-delivering. Constantly. With super high code quality, better test coverage, easier audit prep, and I just explain what I want (technically) at a high level to AI. I still have to code review, and it sucks. Not because the AI needs it, but because auditors need it. I think I can make it faster and easier in the near future with another custom tool.
I actually have time to make architecture diagrams. I have never had enough time to build that level of documentation.
With Fable, I felt the promise of this coming. With Astra, I feel like we finally had the resources to fully professionalize and supply our software team.
We actually need to hire more people next quarter.
I am with the creator of Rails on this. It is a time of joy. Now is the time to become more ambitious.
:shrug: I spend a lot of those token make local models better, resurrecting old hardware to make it more accessible and usable. Frontier models and AI is cool. This is my lived experience doing real systems engineering projects. My real sales manual is: https://stealthgoat.com/deepswe-flash-next.html local models are kind of awesome now and everyone should be using them in addition to frontier models. I will build the stuff I want which includes optimizing local inference engines until they are at my hardware ceiling. I use this capacity in my house to do reversing, security research, and internal OpenClaw type stuff I find useful.
They are all very slow at the moment. Some tasks are taking hours to even days. These are DevOps related tasks in which I create hundreds of servers in an agentic way. It works. But it's slow. That's still annoying.
Gemini 4 is announced be available to Pro users in the Gemini app.
I mean, I heard rumors too that it would be ultra only, but that doesn't seem to be the case, unless you are specifically talking about antigravity?
Realistically, market adoption isn't going to be the result of terminally online people (like me, I confess) following things way too closely on social.
There weren't "rumors" that it would be Ultra only, that's what Google has announced two weeks ago. Happy to see they've went back on that, but I haven't seen Gemini 4 on any apps till now, and I'm a Google AI Pro subscriber.
edit: just a few hours after this comment I got access to Gemini 4 on the Gemini app. Yayy! :)
That's correct, I've misquoted it. Although, if I understood correctly, it got released to all plans together today right? Or did AI Ultra subscribers have access to it since the first announcement?
I went from using GPT exclusively to cancelling all my subscriptions with OpenAI.
'backlash' is underwhelming. They single handedly murdered 5.3 codex which was great by making it slooow and duuumb, and pushed people over to their new ones. 5.6 was a decent candidate but it was so slooooow.
And then double or triple the cost for Plus and Pro by nerfing their effective power? I never looked for an alternative faster. They really screwed the pooch.
Been quite happy with deepseek v4.1 flash, muse 1.3 and haiku 5.5.
"Disastrous" is obviously a loaded term, and it can be debated, but from what I collect, the reasons are
- Dots. People don't understand it. Their presentation didn't explain it properly at all. Additionally, many people thought it's stupid, and that the cutesy mascots are dystopian.
- Voice mode didn't work in their presentation. The demo failed.
- 200$ plan goes 20x -> 10x usage. 500$ plan introduced, 25x usage. The contrast is stark.
- "Ultrafast": 8x pricing on subscription plan (6x API). That's 80$ input, 400$ output, 100$ cache write. Hello?! Additionally, since it generates 8x faster, it consumes your quota 8x faster. Per wall-clock minute, it's 64x more usage spend per minute of generating. People report that on the 500$ plan, their weekly usage is gone in 40-100 minutes.
- People believe gpt-6.1-sol was planned to be released as astra-minor, but OpenAI renamed it at the last possible time, because gpt-6-sol is a disaster (this I can get behind)
Pulled astra 6.1 at the last moment (allegedly) and only released a dirty looper sol 6.1. Also released dots which are unavailable in EU. There may have been more but no one cared.
It is multiple layers of approval, but there is no reason to think they wouldn't approve this. At least not from the limited information I have. The approval is about covering themselves, so if they won't get sued (either because it is legal or they can cover their tracks) they will approve it. There is more than lawsuits they track, but the exact set of things is different for each company and I have no idea what Apple works on.
Thus this could be Apple marketing, or it could be someone independent.
“I have worked over 100 of these kinds of campaigns and never had it come back on the client. I’ve been doing viral marketing and reputation management since 2005. =In the past year I’ve worked for a major entertainment network to magnify a rumor within sports entertainment, as well as damage control on a rumor that came out of an actor being hired on a film before the production company was ready to announce that casting.”
If apple indeed use shit posting, there has to be several layers of subcontractors in between for plausible deniability. Apple's own comms guidelines strictly prohibits pay-per-review, manipulative, "shit posting" style practices.
I understand BBC may not have the technical background to critically assess this approach, but whoever using Eudora (I loved it in 2001 btw) for security should get their credentials removed via King's order. Security through "obsolescence" is no different from security through obscurity - therefore, it doesn't work. Somebody not bothering to look for holes in your software doesn't mean they don't exist - in the age of Claude - I am pretty sure I can destroy your legacy software in minutes.
> in the age of Claude - I am pretty sure I can destroy your legacy software in minutes.
Yes but you'd need to a) know that they use that particular software and b) have a reason to bother destroying it in the first place.
That is really the crux of the idea.
The client the guy in the article uses isn't secure because it has no security vulnerabilities - it is secure because nobody bothers to target Eudora users in general.
Of course as others mentioned, if the target switches from "Eudora users in general" to "that guy in particular" then the situation changes (though the attackers would still need to realize he uses Eudora - assuming this article didn't exist to reveal it anyway :-P).
But aside from that, even "in the age of Claude", i doubt anyone is wasting time and/or tokens scanning the open Internet for all sorts of old vulnerabilities in antique software that (relatively) nobody uses in hopes they catch some random passer-by as there is barely any ROI by doing that compared to taking advantage of vulnerabilities on software that people actually use.
> I understand BBC may not have the technical background...
This is BBC Future - the BBC's tech clickbait publisher - not BBC News.
BBC consists of the non-profit news bureau as well as at least a dozen for-profit clickbait and listicle publishers. BBC's ad-free mandate is only for the UK.
> in the age of Claude - I am pretty sure I can destroy your legacy software in minutes
Yep. One of our PortCos has unrestricted access to Anthropic and GPT models. With whitebox testing, it's trivial to identify vulns in legacy environments. With blackbox testing, it takes some effort but it doable with the right steering.
I would be curious to know what kind of vulnerabilities the latest version of Eudora (7.1.0.9 I think) has. With how old it is (2006), surely there are several critical vulnerabilities, but all I could find when looking online is that an IMAP server or SMTP server can execute arbitrary code, which doesn't seem likely to cause a real problem, because I wouldn't expect Google, Yahoo, Microsoft, etc. to use this trick.
The article links to a vulnerability from 1998, which I expect is already fixed in the versions of Eudora people still use.
I agree it would probably be easy for AI to find exploitable bugs though.
mutt is still being actively developed. Being on command line doesn't make something legacy. Eudora on the other hand had its latest version out 2 decades ago - it lacks even the most basic security updates (starting from TLS, but many more).
reply