It's really bizarre at this point. I'm okay with things like having one-click options for simple replies like "That time works for me" (Google Messages on Android is hilariously bad at these but it's at least useful occasionally). I'm not okay with it suggesting a whole point-by-point response to someone else.
Today I had an email from a colleague where it had a suggested reply so large it didn't fit in the preview box. The response was the usual LLM "sounds good but doesn't say anything" prose. It's not just unhelpful — it's a waste of everyone's time!
One could argue the “That time works for me” prompts are also a waste of time, especially when only occasionally useful. Do we really need a button to type 5 words on an occasional basis?
I’ve used the button before, but I think my life would be exactly the same without it.
I have to wonder if they just give us the illusion of being useful.
Just had an email asking if I could meet "tomorrow at 5:00 PM, or would 6:30 work better?". The suggestions were "I will be there at 6", "Tonight works" or "Either time works", only one of which is even valid! Maybe for every time it saves me a few seconds, there's at least one where I have to read them all and realize none fit before writing what I would have done without the quick replies.
Even "either time works" is only half-valid! If the other party has already declared their openness to either option, proper etiquette is to just select one so you can both move on with your lives.
That's actually a use case that I imagine could work well, if done well. Especially in fully integrated systems like GMail or our corporate Exchange, when the LLM can access enough data to produce meaningful suggestions.
IMHO the UX problem is, as the article points out in so many words, shoving AI slop down our collective throats as if we were geese waiting to be fattened.
If you’re getting so much email you can’t type five words, a button with wrong suggestions isn’t the solution either. Create your own text snippets with text expansion software and use those. They’ll always be correct and will truly save you time.
let me try a different tack - 99% of the value is in the affirmation of the response and the demnonstration of responsiveness, not the actual content value of the 5 words. you could send a thumbs up emoji, you could send a ".", just the "seen" and "acked" value is the value of that 5 word button
Which again underscores the question of why we need an auto-generated response when a thumbs up will do.
Of course a reply is valuable, but are those options really saving any real time vs a quick single word acknowledgement, which is often faster than even reading and considering the options?
Are we shocked, really? We have evidence from discovery of them actively making search results worse to pad query volume. Of course they’re using enabled-by-default, run-without-asking AI features to pad Gemini usage.
The Valley is tripping over themselves to convince the world fancy autocomplete is worth 800 billion.
The economic tension of these "run by default" AIs is quite hilarious once you see it.
On the one hand, the relevant KPIs of whoever is driving this product needs to be able to show AI usage is increasing, because AI usage is obviously the Platonic embodiment of goodness [1].
On the other hand, these things are expensive, so while it's mandatory that Google searches stuff these things in our faces, they are also horribly underprovisioned. If my only exposure to AI was the various search engine popups or the other free AIs, or even the bullet-point AIs that I'm nominally paying for but not really, like in Office, I would also have a pretty negative view of AI. I use DuckDuckGo more than Google but whatever model they may nominally be using to power their search result summarizer, it is de facto at least two years behind the state of the art in a very fast moving industry. It frequently gets things exactly backwards and is clearly leaning on its internal model a lot more than the links it has supposedly read, and clearly has a thinking budget of "indistinguishable from zero", and I don't know what kind of summarized web page content is being fed to it but it must be getting brutally dismembered in whatever summary is being fed to the AI.
The debate about how useful AI you pay for may rage on, but at least at this point in 2026, I'd say the AI you can get for free is every bad thing anyone says it is.
[1]: I believe there is a lot of useful things current AI can do, but there is no level of quality AI can ever reach in which AI usage for the sake of using AI will ever be a terminal good. Honestly any manager, whether they be a line manager or a CEO of a multinational company, that has ever pushed that in any capacity, should be fired for demonstrating gross incompetence for that position. It's "second or third week of Econ 101" or so that you learn about why it's never a good idea to just open the checkbook and spend an unbounded amount of money on something, and nothing you'll learn further down the line will ever contradict that.
It’s not bizarre once you realize it raised some team’s engagement metrics (like # of emails sent) by 2%, a result they were unable to achieve otherwise, so they shipped it and celebrated a win
If people dislike subscription-based games, companies will adapt by making non-subscription games designed with end-of-service in mind. It only creates an incentive as much as people are willing to pay for the subscription.
> It's entertainment. It's ok for entertainment to end
Sometimes it ends right after you bought it with no way of knowing it would, or before you bought it. Not everyone gets 1000+ hours out of a title, sometimes the day you install they announce that the servers are going down forever.
I'd rather discuss actual problems people had with specific titles. Hypothetical edge cases always turn into corner cases during cross-examination. If we think about people who bought a game closing on day one we need to think about the people who shop at stores with a no-return policy. I'm not sure who's problem that should be.
>I'd rather discuss actual problems people had with specific titles
Here is one - I never got to play The Crew which got closed in 2024. But I would certainly like to do that now and some time later (10, 20 or even 50 years later). The same goes for Diablo 3 and 4, which did not shut down yet but they might. I just want to be sure that I will be able to play a game in the future. I do replay and reevaluate games, so for me it's important to keep them available all the time. Especially if I paid money for them.
> you will just get a limited and nerfed Claude Code slash command /mythos-security-audit or something.
Unless it's so expensive that we can't realistically use it for anything, I wouldn't complain about getting at least that. I would also rather have the actual model, but that's a useful application of it (and I'm probably not going to afford using it for much more).
Price discrimination is I think fine and reasonable so long if you can drum up the cash you can use it how you want within their ToS.
Although mental safety gymnastics aside, getting the most amount of intelligence for the cheapest amount of cost to normal people seems like the most ethical thing a big lab could do.
Going around and granting different tiers of intelligence to different insiders, friends, or companies is majorly problematic long-term.
Heck right now, the tokens you buy today for “Opus 4.8”, no one even knows or believes will be the same “Opus 4.8” just 3 days from now.
/security-review already exists so I don't think it would be crazy to have a /mythos-security-review as more thourough command as well. I think it's more likely it is going to be released at some point to the general public though - although the the pricing might make it quite unattractive.
It seems like a fine question to me. If the question is "logically inconsistent" (IMO it's more that it's vague if you don't say why you're going there), then we want a model to respond with a request asking for clarification that resolves the inconsistency to generate a correct answer, or an answer that outlines the different cases. Some models even fail when you say that you need to wash your car in the prompt.
Yeah I guess it being vague is more what I meant. But even if you told AI you need to wash the car, then why are you asking AI in the first place whether you should walk there or drive there. The question just doesn't make too much sense to me, doesn't look like it makes sense to the AI's either.
A lot of people care about Sonnet and Haiku, and many of us aren't allowed to use Chinese models for our work (or it's not feasible to self-host them).
reply