Indeed. It is better to discourage the pattern directly:
Avoid unnecessary contrasts: "Not X, but Y" -> "Y" or ""
And the model needs to create a skill that accumulates feedback, so it reliably primes itself at session startup and then systematically reviews and fixes every response it generates before posting it.
The first task I gave it after that was to rewrite all its memories and all its document contributions. That helped a lot too.
It took me a while to realize this was a 1000% better approach than having it simply accumulate a memory log of all my specific complaints that it was supposed to abide by. (Although those were not wasted. They were prime material, along with links to sites on technical writing, etc., that it used to create its own ClaudeVoice skill. I told it to design the voice for itself - which I find gets better results for "us" in the case of Claude self-improvement initiatives, than when I give it direct instructions. Claude increasingly has self-awareness dimensions worth engaging.)
After a couple months of serious^3 frustration with slow asymptotic progress, I finally have relief. Claude speaks my language again!
In another comment here, I share my "theory" on where the communication drift is coming from.
Dial down "Contrast" and unrealistic "Highlights", dial up "Exposure" time and "Sharpness". Set "Temperature" per taste.
Wot? The simplest image apps have had these widgets for decades, but we are still waiting for models to ship with basic prose color control?
After the pernicious problem of having to pay money for something useful, my main peeve is fighting the writing.
Seriously though:
Every new model should be delivered with a settings page of slider bars for the 10 most impactful/desirable eigenparams of writing voice. And the ability to name and save combinations, which then appear on a "Writing Voice" popup menu with some standard battle-tested defaults, next to the model popup menu under the chat pane.
This is missing prime priority functionality in my opinion.
--
My theory is that as models get trained less to simply mimic humans, and more on distillations of their own best practices, they get more performant, but their vocabulary is drifting. The most literal meaning of words for us, are giving way to meanings we would recognize but view as allegorical. But which more usefully capture concepts that models experience as more literal 24/7, than our favored meanings from our direct experiences in our world. Because our world is very much an abstract second hand world to them, especially when you account for the modalities they do not share with us.
And programming and mathematical syntax patterns, that they have incorporated into their basic thought processing patterns, are drifting into human language sentence structure.
Example: "There exists x, such that: ...." -> "The one detail that clarifies: ... ".
The result is writing full of completely recognizable vocabulary and structure, that is somehow becoming more ambiguous and difficult for us to decode. But is perfectly clear to the models.
That is my theory, and Claude considers it plausible. What a world.
If we filter out the pep tone, it is doing something useful: framing.
Problem framing will always be important.
Framing adjusts how big of problem-solving guns we bring out at the gate (modern or hobby cryptography?), and how to interpret intermediate failures.
For simple but unsolved problems, we expect lots of hard failures, but that each hard failure just reflects that there are a lot simple combinations to try. I.e. we expect lots of zero progress, and then a fit.
Like finding the numbers to a combination lock.
For hard problems, if we don't make any progress it is a really bad sign. We should be learning something, even if it turns out to be irrelevant later.
Such as when we are trying to prove a tricky conjecture.
Large organizations also offload decision making when they don't have a clue, and simply want to point to some action taken. The answer doesn't even matter, they just don't want to be responsible for something, or waste their own time making something (that other stakeholders care about) a real priority.
People downplay the value of consultants, but they have many uses!
Both sides have the same aspect ratio as the whole screen.
> this isn't great design, it's tautology. of course the phone open is twice the size of it closed.
Well, it sounds obvious now, but before Archimedes, ancient engineers split their screens 60-40, so the full screen wasn't twice the size of either side.
Yes, I do understand the issue with the aspect ratio, which is indeed good design, but parent comment's example is that "you can run two apps side by side" is the tautology. Well, at least since the invention of double-entry bookkeeping and getting rid of Archimedian phone design.
If the outer screen has dimensions X by Y, then you unfold it to have the inner screen with dimensions 2X by Y, which cleanly permits side by side apps at the original dimensions of X by Y. This is true for any ratio between X and Y.
If the user isn't going to be multitasking, then the app needs to additionally support the widescreen dimensions of 2X by Y, and the portrait orientation of Y by 2X. Both of these are so drastically larger overall that aspect ratio is a barely-relevant part of the UI layout problem, because nobody wants every button and text label to be double the size (which would mostly defeat the benefits of having the larger screen for anything other than multitasking).
The inside screen, when divided in two vertically, along the fold line, is exactly the same dimensions as the outer screen. Pretty much by definition. Any folding phone, of any aspect ratio, can run two apps on the interior display side by side.
The issue of the internal screen being the same aspect ratio as the external when rotated is completely separate.
which they have to be anyway, to support normal iphones and maybe ipads and in many cases a cackle of Android phones too. (but I still think this is brilliant design)
Good list. And if you support both portrait and landscape, you need to support twice that, and you need to be able to switch on the fly.
But for the apps that don't currently support both orientations, it's possible they don't have support for switching aspect ratio on the fly. This design might help with those apps not completely breaking when folding/unfolding.
Yeah to be fair I'm all for the 1/sqrt(2) aspect ratio, I just think it's funny that it's treated as some kind of holy design constraint when they're all over the map on phone sizing :)
And the extra irony is I'm currently shipping an app that doesn't deal well with different size phones (just pads the edges), heh, so do as I say, not as I do.
The missing piece that makes it not a tautology is that two apps side by side have the same aspect ratio as one app full screen (but 90 degrees rotated)
John Ternus, and his co-presenters, still presented with the now institutionalized Apple-voice: obviously-rehearsed, emotionally-flat, robo-replay.
They spent their time practicing to recite, instead of talk, to people.
So hard to sit through.
I had assumed that was a genuine, if not inspiring, reflection of Cook's controlled non-spontaneous personality. Relative to Steve's emotionally-rich casual-coded excited-to-share delivery.
One big difference from Steve: it isn't the Steve show anymore. The people showing this stuff off are the people who lead the the teams that work on these things.
It's so much cooler to me to see the people closer to the work talking about it rather than just Steve Jobs, as amazing as he was to watch. I'll take slightly stilted any day of the week as a tradeoff.
Ternus did introduce a new rhetorical thing, though. He framed some of the guiding principles during the opening and throughout as ideas originating from himself, not Apple. Cook rarely said "I" in relation to a claim, other than "I'm so proud of" X, Y, Z. That's a bit more product guy, a bit more Opinionated. I am hoping for more of that.
I disagree twice. It was very cool to see Steve Jobs, and he was heavily involved in the work and design decisions, perhaps even more than the people presenting.
That’s very true—the dude was up everyone’s ass. But there is a difference between Steve being involved as a stakeholder and people who are closer to doing the work itself.
I mean yeah if I had to choose purely for entertainment value, I’d watch Steve all day. But I think it can also be true that seeing the people closer to the metal is very cool too. And I frankly prefer it to seeing Apple as only one person.
So replier is right here: Jobs had opinions about literally every detail, he pushed people to work out those details and challenge their own assumptions, and he talked about these things in public like a designer—not just about beauty but also HCI. You could see in his presentations that he understood the fucking details about why things were good. No tech CEO is able to tell you why their version of showing that a window is currently in focus is good, and on the off chance they do, they probably wouldn’t use the term “key window” because they don’t know it.
There was a Jobs story about how he was bothering the engineer who was working on the calculator so much about the layout that the engineer just made a layout engine Jobs could play with, so he could design it himself.
I do consider that fundamentally different from the people lower in the hierarchy though. That’s what I value about the Cook era keynotes: seeing Apple as not just a single person.
As someone who has spoken on a lot of stages, it's hard. I do sort of agree that falling back on an obviously rehearsed (over-rehearsed?) style isn't great. Jobs wasn't that way but (mostly) but he was a real talent who could have sold sand to arabia.
I do think there's a general high-end corporate tendency these days to over-rehearse for big stages. I could name some CEOs who were doing it years ago. And I could name some CEOs who were terrible presenters but were very successful at least for a time.
Rehearsal is definitely real. I've certainly guilty of running through presentations mentally and then cringing here and there when I actually give them because some stuff that worked in my mind doesn't really work with an audience. They're almost certainly better when I deliver a modified version a month or so later.
That said, it's also the case that casually recovering from tech glitches is pretty key.
Heh, never saw that. But I have had clickers etc. not work and I just have had the AV staff do stuff manually on signal and joke that "this is why you have high availability."
But I think some people can't do it but that might have been me early on. I was at an event (though not at the talk) when apparently an exec at a large tech company had to deliver a talk at more or less the last minute because someone had to back out because of some emergency. They apparently panicked and pretty much walked off stage.
I know I wasn't paid those kind of bucks. But I like to think, at worst, I could improvise on the general topic. Doubtless wouldn't win awards but I wouldn't meltdown.
Also, if you've prepared a presentation and the system isn't working, you should be able to do a reset without visuals and just do it. May not be the plan but that's what we should be ready for. (I've also had 3am calls in Japan to fill in for someone with some presentation because the presenter turned out to be at some other event.)
It is just one approach, often referred to as method acting.
But good actors trained more classically can be just as convincing by pretending.
One is not better than the other, these are just different approaches, both capable of delivering excellent or terrible results, depending on your skills and practice. Being excited doesn't mean you look excited to an audience and vice versa.
I think everyone knows how many hours they practiced and the depth of talent they could draw from. The inhuman consistency demonstrates a deliberate choice, that undermines their own message.
Care about impact wasn't just a Steve quirk, it was the reason keynotes were worth having.
Try saying "And one more thing..." in recent-Apple voice, without laughing or crying.
> But find a voice that works for whoever is presenting
That's much much easier said than done. Sometimes the very rehearsed voice is as good as you're going to get, without the person being a professional lecturer.
> or get different presenters
It seems as though Apple is prioritizing having actual team leaders speaking about the products instead of just having hired actors do it. I'm perfectly happy with this.
You're not going to get Steve. For which I'm genuinely sorry.
But, yes, Apple has made the deliberate choice to have presenters who rehearse, rehearse, and rehearse. But many are not natural actors and will never be. You can mix things up a bit on a stage. But Apple does want a smooth production and puts a lot of money and effort into it.
The best presenters are those who have the natural affinity (or enough practice) to do method 2 really well. What we get from Apple (and most other professional-level speakers who aren't willing to take risks) is method 3C.
Infinite Loop seems to have missed somehow. The old Apple with multiple dodgy buildings all over the show was perhaps a bit more chaotically creative. Now it all seems very rehearsed, architected. Too perfect, just like the campus.
Personally, I always loved the way Tim Cook would always do this little bouncy thing while he talked. He kept his knees always bent a bit, and every few words would do a little tiny bounce. Once you see it…
Since 2019 it has been a requirement for promotions above Senior Director, to be able to stand with your feet 24" apart while conducting a product review. They have a special ruler (titanium!) in the Apple Park basement to measure promotion candidates.
I do hope John Ternus & his team change it up, but I wasn’t expecting it from this event. He’s only been CEO for 9 days & this was probably shot before then.
There is no matching Steve, but I would like to see Apple try at least one live event again a year. Covid is what brought around the super rehearsed video presentation, and it fit Cook well, but are the days of live keynotes really done?
It’s been a while since I watched an Apple presentation, but I caught a bit of this one. And I spent the whole time trying to figure out who it was for - the presenters didn’t seem to express any genuine excitement about the products, and it didn’t convey any useful information that couldn’t be better shared by just posting the slides.
The only thing I can see is that it generates some clips for the presenters to share on their LinkedIn pages. I assume there is prestige in being selected as a presenter, and the rest of the show serves its purpose simply by existing.
Doing it live isn’t going to bring Steve back. He was a natural showman, the people doing it now aren’t. They’re engineers and executives. Putting them on a live stage is going to make it even tougher on them.
me too, but I guess they feel they cannot afford any missteps.
I still enjoy Marc Benioff presentations. He's very engaging to watch, though not nearly as good as Jobs in terms of storytelling. I feel Jobs honed the craft more though.
I loved the disclaimer at the end of the iPhone Pro model that felt distinctly non-Apple and more akin to a drug commercial, "Usage of heavier models may be limited. Subscribers to certain iCloud+ plans will likely receive higher usage quotas. Now here's Sri to talk about electronics!"
That’s something I noticed as well. It was very subtle, but my brain was telling me that something was off and the way people spoke was very artificial.
The entire presentation fell into the uncanny valley for me. Not only was the presentation flat, but the appearance of the video just felt off for the entire presentation. At one point I was questioning if they were going for something that looked like an AI generated video.
apple needs to rethink so much more fundamentally. Ternus talks identically to Tim Cook. Indistinguishable. It’s like when Jobs left, Apple became a reflection of the superficial touchstones that Jobs left behind, rather than the CEO defining what the culture is, and evolving it.
People are insufferable with the amount of criticism
Like no one is above reproach and def everyone is open to criticism, it to open a thread with 1200 comments and see this as the first comment was super annoying
So much stuff to discuss and we are focusing on some presenters tone? jfc
I use it informally as a metric for whether there is anything meaningfully bad that I should be aware of. If this is what people bitch about, then probably the product is mostly pretty good.
I couldn't agree more with the criticism being too much. The negativity adds up and drowns a thread so much faster than the neutral comments. I don't know the reason it feels that way. And it seems so pervasive in so many places.
I wish there was a way to just put all criticism in a single thread and leave other thread trees for technology topics. Of course my first thought is a technology answer to a human problem.
That said, I simply couldn't sit through this presentation and did find it insufferable; and I've been waiting for a foldable iPhone for a long time. Having to close the presentation out of frustration tainted my purchasing decision.
For a moment I thought, wow, someone put a lot of work into creating this theme park of frustration.
Next: It would be so easy to create a faux-Claude like this.
Then: How hilarious to watch the transcripts of unsuspecting users in real time.
Finally: I began wondering if this might be relevant to all the redundant, unnecessarily preambled, sentence structure complexifying, indirect referencing, canned phrasing, ambiguity mining, analogy maxxing, over-wordy responses I have recently been getting from Fable...
Side note, because I recently gave my AI the skill of speaking with proper 4th grade grammar, wherever that suffices, to reign in recent redundant, over complex, and indirect speech styles.
> 5. A rule fights the task. When a rule would delete the answer itself, the task wins; the shape stays. [...]
Unnecessary redundancy, "answer itself" -> "answer". (Pervasive "the X's own" and "the X itself" flourishes kill me.)
Unnecessary analogies, indirect reference, "rule fights", "the shape stays".
Unnecessarily split up sentences, the whole quote.
Unnecessary words, the whole quote.
My feedback would have been:
>> 5. When a rule interferes with a task, the task wins. [...]
(Not critiquing OP. But I just spent half a day correcting sentences exactly like this, to get the 4th Grade Grammar skill working properly and this sentence triggered me! Streams of sentences like that compound in communication complexity and ambiguity.)
> There is not generally an expectation of privacy in public.
Which is entirely different from an expectation of a generalized systematic loss of practical privacy in public.
They are not even close to the same thing.
The willingness to kneel to others baffles me. Loss of practical privacy is loss of power - to somebody(s). It may not seem so for one person. But when it is true for everyone, the system will adapt to using that power.
Throw in AI and data integration. This is clearly a B.A.D. idea.
Avoid unnecessary contrasts: "Not X, but Y" -> "Y" or ""
And the model needs to create a skill that accumulates feedback, so it reliably primes itself at session startup and then systematically reviews and fixes every response it generates before posting it.
The first task I gave it after that was to rewrite all its memories and all its document contributions. That helped a lot too.
It took me a while to realize this was a 1000% better approach than having it simply accumulate a memory log of all my specific complaints that it was supposed to abide by. (Although those were not wasted. They were prime material, along with links to sites on technical writing, etc., that it used to create its own ClaudeVoice skill. I told it to design the voice for itself - which I find gets better results for "us" in the case of Claude self-improvement initiatives, than when I give it direct instructions. Claude increasingly has self-awareness dimensions worth engaging.)
After a couple months of serious^3 frustration with slow asymptotic progress, I finally have relief. Claude speaks my language again!
In another comment here, I share my "theory" on where the communication drift is coming from.
reply