Reimagining the mouse pointer for the AI era(deepmind.google) |
Reimagining the mouse pointer for the AI era(deepmind.google) |
Assuming that today the most efficient way for human to transfer information to machines is via voice. Assuming for machines to convey rich information to humans that's by printing html.
Then a combination of screen + eye tracking + voice is all you need. The mouse doesn't make sense anymore.
Wait…it's May. Ugh, I'm so confused. :spiral eyes emoji:
Unless of course, their AI gets the same special privileges as the gpu in accessing drm content, and everything else is still locked out.
They have so many great software engineers but unable to use them to speed up coding AI research. Hopefully with Sergey's focus it will get better.
This cursor thing is just another experiment nobody cares about.
Would be tiresome though to hold hand out all day - but good for mobile and handwriting/drawing. Need zero latency.
Furthermore, the mousepad could become the magnetic sensor and not the screen in order to rest the palm. The nail bands then become the equivalent of the mouse so it's a hall effect mouse. But could the pad detect finger twitch for the buttons, though?
1. select text
2. dictate action
Feels very similar to Helix's select text and act on it.
I think text selection could also be voice controlled (with a modal voice input), so one could say: "select sentence, action mode, copy and paste it in my list and remove duplicates"
We couldn't quite track you well enough before. So we're fixing that under the guise of "AI powered capabilities."
Horizontal dragging with a mouse is actually really hard. Nobody's going to use it like that.
Your arm can easily move your hand and cursor up/down by pivoting your shoulder, but there's no mechanism for left/right movement. It's always an arc.
Or put another way: selection will be a lot slower and more tedious than the demo.
But Google is a very ill positioned candidate for such OS. I would rather trust Apple and local-first on-device models.
I'm mostly using my system to make comments on long AI-generated documents (especially design documents). I find it works well to have the AI generate something, and then I read through it, making comments along the way.
You can get pretty far just repeating the things you see... "I'm reading [heading] and [comments]". But I do find some use in selecting content and saying "I don't agree with this" or whatever else.
The result is just an augmented message. It looks like:
<transcript>
Let's see what we've got here.
<selection doc="proposal.md" location="paragraph 3">
The system already...
</selection>
No, I don't like how this is approaching the problem, ...
</transcript>
It makes reviewing feel pretty relaxed and natural. I've played around with similar note taking systems, which I think could be great for studying in school, but haven't had the focus on that particular problem to take it very far.
But I think the best thing really is giving the agent a richer understanding of what the user is experiencing and doing and just creating a rich representation of that. The keywords can be useful, but almost only as checkpoints: a keyword can identify the moment to take the transcript and package it up and deliver it.
One difference perhaps in design motivation: I have really embraced long latency interactions. I use ChatGPT with extended thinking by default, and just suck it up when the answer didn't really require thinking. I deliver 10 points of feedback at once instead of little by little. (Often halfway through I explicitly contradict myself, because I'm thinking out loud and my ideas are developing.) I just don't stress out about latency or feedback, and so low-latency but lower-intelligence interactions don't do it for me (such as ChatGPT's advanced voice mode, or probably Thinking Machine's work). I think this focus is in part a value statement: I'm trying to do higher quality work, not faster work.
you select text in vscode, and write a comment, and the llm gets both
Perhaps a text box and file upload isn’t the perfect interface for every use case but it is versatile which is a huge barrier to overcome.
font-feature-settings: "ss02" on;I also don't think people want to constantly talk to their computers.
People don’t. Tech companies think they do.
Nightmares are dreams as well and this is a nightmare like Windows Recall.
Technically wonderful though.
This has a good utility.
At some point I fully expect eye tracking (or attention tracking) to be common enough to be a first-class input method.
There's a reason chairs are still around. They are +2000 years old. Its still waiting to be replaced.
One should be extremely skeptical of claims of replacing tech that has been around for a very long time.
Bullshit!
You don't need any new metaphors to support such (questionable) flows - at all.
Swipes instead of selection rectangles are annoying - you don't see the traces of the swipes on these demo gifs! So, you've effectively "selected" something - but you have to keep in mind WHAT you selected.
Total ridiculous bullshit.
This reads like an April Fools joke. Even the title sounds like satire.
I'm hoping for a const-reference joke.
Aaaaand now I can't remember the name of it
Only when you live alone might you be comfortable constantly speaking with your devices. Only if your life if perfectly predefined can you let your fridge order the same food that just gone stale or has been eaten. And only when you are young and healthy and not in any way differing from the "standard" would you be capable of working like these "researchers" imagine you to.
I'm not that person. I'm constantly failing at doing "triple-finger-taps" whenever I'm in need of one. I have a smartwatch with pedestrian navigation and never bothered to remember which vibration pattern means which turn. I don't configure different vibration patterns for different callers on the phone. I have a folding phone, but I almost never do side-by-side windows and when I do, I need to find out how to do that first -- and then how to leave that mode without losing my mind. I almost never use AI features on my phone not because I don't want to, but because I never remember how to activate them. I don't re-configure my gadgets to "fit my mood". I hate recommendations like "you like X, here's Y, it's the same!" I hate that I can't rest my mouse cursor on websites anymore without selecting something actionable, moving, animating or autoplaying.
All of the examples on the linked page are workflows I would never do this way. I won't be talking to my shopping list to double the ingredients. I won't be drawing gestures with my mouse on a document to activate a voice command. I won't use voice commands in general because as it turns out, I'm not capable of bringing out a complete coherent sentence without pausing and/or changing my mind and/or realizing I'm wrong once.
I appreciate those demos for the progress they are showing. It's impressive and astonishing to see restaurants getting extracted from videos or pictures getting expanded or text edited better than I ever could. It's all modern-day magic in a way. One thing it all isn't is a product. We don't have those anymore -- all we get are gimmicks. We don't do common interfaces anymore either, we are separating people in Google/Apple/Xiaomi camps.
And most importantly we don't use that technology for good except for a bunch of people writing e-mails all day, doing shopping lists and booking one of top restaurants in Tokyo for the same evening on a whim. We are long overdue for a remake of "American Psycho", but this time it will be a documentary instead of a satire.
>Sousveillance (/suːˈveɪləns/ soo-VAY-lənss) is the recording of an activity by a member of the public, rather than a person or organisation in authority, typically by way of small wearable or portable personal technologies.[14] The term, coined by Steve Mann,[15] stems from the contrasting French words sur, meaning "above", and sous, meaning "below", i.e. "surveillance" denotes the "eye in the sky" watching from above, whereas "sousveillance" denotes bringing the means of observation down to human level, either physically (by mounting cameras on people rather than on buildings) or hierarchically (with ordinary people observing, rather than by higher authorities or by architectural means).[16][17][23]
Anything with voice controls for routine use is a pretty tough sell. Doing this when you're not completely alone would be annoying to everyone around you.
Most of their examples seem like they could have been done with a right click drop down menu so they don't really need to "re-invent the mouse pointer".
So is this thing talking to Google's servers all the time for the AI integration? So it won't work if you're not connected to the internet? Privacy concerns are obvious; now Google wants to have an AI watching literally everything you do on your computer?
Does it cost the user anything for the LLM use? If it's free will it stay free forever? That's quite a lot to give away if they're expecting people to use it to change a single word like in one of their examples. I guess they're expecting to make the money back by gathering data about literally everything you do on your computer.
There might be a killer app for AI integration with personal computers that has yet to be invented, but this doesn't look like it.
What's being delivered now is, an agent running on someone else's computer, copying your data to someone else's database, with zero responsibility, or mandate to protect that data and not share with with anyone else (in fact, they almost always promise to share it with their thousand partners), offering suggestions and preferences based on someone else's so-called recommendations, influenced by paying the agent's operators, and increasing pressure to make using someone else's computers + agents the only way to interact with other people and systems.
There is no doubt that LLM's can do amazing things, but the current environment seems to make it nearly impossible to do anything with them that doesn't let someone else inspect, influence, and even restrict everything you are doing with with these systems.
If we're going to have AI regulation, this is where to start. If a company's AI service acts for a user, the company has non-disclaimable financial responsibility for anything that goes wrong. There's an area of law called "agency", which covers the liability of an employer for the actions of its employees. The law of agency should apply to AI agents. One court already did that. An airline AI gave wrong but reasonable sounding advice on fares, a customer made a decision based on that advice, and the court held that the AI's advice was binding on the company, even though it cost the company money.
This is something lawyers and politicians can understand, because there's settled law on this for human agents.
The hard reality is that you are still responsible for all of these things. If anything goes wrong at all, you are liable. Might not be devastating if it's just your shopping list or your photos mangled, but with taxes or bills? Even if the agent is running completely locally in your home, you still won't trust it fully if your livelihood depended on it.
The killer app is only possible if software is fully reliable, which we all know is not the case. Software is just that: software, it still has bugs, undefined behaviour etc. Agents are the same, they just break in different way and fixing them might be even more difficult.
Bottom line: you will always be liable for things happening in your name and we've been sold a fairy tale a very long time ago.
I guess what I'm saying is - we've always had this problem.
But I don't think the voice problem is surmountable. I closed their image editing demo when I saw it required a mic.
It would be appealing as a Spotlight-like text pop-up interface where you type instructions, which would work in social/office environments, but that might only appeal to power users.
But if it's going to require phoning home to some Google/OpenAI/whoever then forget it. I don't want a constant connection to my OS from one of these companies.
https://www.youtube.com/watch?v=46EopD_2K_4
>We present a general-purpose implementation of Grossman and Balakrishnan's Bubble Cursor [broken link] the fastest general pointing facilitation technique in the literature. Our implementation functions with any application on the Windows 7 desktop. Our implementation functions across this infinite range of applications by analyzing pixels and by leveraging human corrections when it fails.
Transcript:
>We present the general purpose implementation of the bubble cursor. The bubble cursor is an area cursor that expands to ensure that the nearest target is always selected. Our implementation functions on the Windows 7 desktop and any application for that platform. The bubble cursor was invented in 2005 by Grossman and Balakrishnan. However a general purpose implementation of this cursor one that works with any application on a desktop has not been deployed or evaluated. In fact the bubble cursor is representative of a large body of target aware techniques that remain difficult to deploy in practice. This is because techniques like the bubble cursor require knowledge of the locations and sizes of targets in an interface. [...]
https://www.dgp.toronto.edu/~ravin/papers/chi2005_bubblecurs...
>The Bubble Cursor: Enhancing Target Acquisition by Dynamic Resizing of the Cursor’s Activation Area
>Tovi Grossman, Ravin Balakrishnan; Department of Computer Science; University of Toronto
I've written more about Morgan Dixon's work on Prefab (pre-LLM pattern recognition, which is much more relevent with LLMs now).
https://news.ycombinator.com/item?id=11520967
https://news.ycombinator.com/item?id=14182061
What if I am going on vacation next week? What if I need extra milk for a dinner I am planning? What if my kid puts the milk in the fridge sideways and it no longer detects it?
"Easy fixes" to easy problems never work because they add mental load to tasks we already manage capably. Yes we no longer have to think about buying milk when it gets low, which was a stable pattern. But we replace it with a nondeterministic "milk state" that we need to be constantly vigilant about and manually adjust any time our routines are altered - exactly when we don't want to stack on more overhead.
AI is discretely useful, tremendously so, but big tech loves to default to umbrella solutions before there is a rich context to reasonably support it. The real world is messy.
General AI product tip: show your tool fixing a messy problem not a happy path problem. That's where AI is impactful!
I think its brilliant UX.
Wouldn't SilentWhisper do just as good a job?
First things that came to mind:
- facial hair
- getting people to learn to make bigger mouth movements and not mumble
- we're constantly self-correcting our speech as we hear our voice. This removes the feedback loop.
- non english languages (god forbid bilingualism)
- camera angles and head movement
I dunno how I can express this best, but I found out a very long time ago that my problem with voice input wasn't that it wasn't good enough. My problem with voice input is that I don't want it. I am very happy for people who use these tools that they exist. I will not be them. Yes I am sure.
And yes, I know SuperWhisper can run offline, but it is a notable benefit that versus many modern speech recognition tools my keyboard does not require an always-active Internet connection, a subscription payment, or several teraflops of compute power.
I am not a flat-out luddite. I do use LLMs in some capacity, for whatever it is worth. Ethical issues or not, they are useful and probably here to stay. But my God, there are so many ways in which I am very happy to be "left behind".
Right-click menus can get cumbersome. I've seen a lot of software that suffers from function bloat - not that the functions don't work, or don't play well together, but that the user interface becomes too overwhelming for users as the number of available actions explodes. This is particularly tough for new users.
This is where voice controls could shine: as we interact with computers in more and more complex ways, we need a way to specify our desires simply and easily. And if we can't do so easily, the software has to remain simple to be usable.
I'm surprised sub-vocal HCI isn't better developed by now. Perhaps because of this stuff coming out it will be.
Humans speaking to one another is literally telepathy: I'm putting my thoughts in your head, with lots of ambiguity and noise, of course.
With better sub-vocal tech we can control our devices without bothering each other.
https://www.media.mit.edu/articles/exclusive-startup-lets-yo...
I assume they're using on-device Gemini Nano: https://developer.android.com/ai/gemini-nano
It's a cool idea for the future when we have reliable EEG headsets or Neuralink or whatever though.
I'll talk to a computer, even in an office setting, if it adds enough value. But it's got to be a lot of value. Handsfree while driving is great, Iron Man talking to Jarvis while he's flying around makes sense. Many of us here are developers, engineers, or scientists, and our work has already been co-optimized with mouse and keyboard and whatever software we're in.
But when the software is less well-developed, or when it's not just dealing with technical data dumps, I imagine that a voice interface might be more useful.
So I think this idea has legs. But a successful implementation might also well be decades out.
Depends on how many hands you've recently broken?
Reads like the argument against cell phones where don't have a cabinet around you...
I'd go and find a small meeting room or conference call booth in the office and take it there.
Essentially, a cabinet.
In fact, when humans happen to order other humans, it's typically done in writing.
(And if it's an abstract entity like a file, it might not even be possible to describe it, short of rattling off the entire file path)
Sometimes I go to a different page to take a screenshot and other times I'm browsing for a file, and other times I'm highlighting some log lines. Cursor did this well, with selecting text in the terminal auto-focusing the Cursor agent textbox so you could talk to the agent and then select some text and you didn't have to re-select the original agent textbox again. The agent is a top-level function in that system not "just another app I have to switch to" to take my context with.
I have some small amount of bias because I've always felt input-constrained on computers. I have to move my hands to go places and that's exasperating. I've tried head tracking, had a vim pedal for a while, and used tiling WMs, and things like this to aid but while my vim-fu is pretty good and I function inside things very well with it, my cross-application interface isn't.
In the end, perhaps we all have our home offices with our Apple Vision Pros and we talk to them like this to maneouvre faster through our machines and get our ideas into them.
Cool research. I wonder what we'll end up with.
Why not constrain your computing? It will require some programming chops, but you can note down your common tasks, figure out where actual input are required, and automate the rest.
The second demo seems to be a wash: there's no time saved in saying "move this" versus "move crab". And an app-specific contextual menu would probably be faster.
The third demo doesn't seem to warrant the use of a pointer at all, since there is only one way to interpret the prompt.
None of this means that this approach will not be successful, but there's a reason why so many attempts to revolutionize user interfaces ended up going nowhere. Talking to your computer was always supposed to be the future, but in practice, it's slower and more finicky than typing.
In fact, the only new UI paradigm of the past 28+ years appears to have been touchscreens and swipe gestures on phones. But they are a matter of necessity. No one wants to finger-paint on a desktop screen.
For you and me, who have used keyboard in our lives for more than 1,000 or even 10,000 hours.
There was a brief period when typing slowed people down because they could write the same information down with pen&paper, and that period eventually passed.
and that's aside from the obvious privacy problems.
And being able to take photos/videos with the glasses (like the Meta ones nowadays) is really useful with my kid because he often does funny or cute stuff and I don't have time to pull my phone out to take a video/photo of it. I guess it could be useful for video calls too so my parents can see him.
But I just don't see anyone sitting in an office, or even at home, talking to their computer. It's really only useful for hands-free settings like when you are driving, or in the kitchen etc.
It reminds me of Microsoft Recall in the sense that some portion of the screen is going to be continuously transmitted outside of the users control.
What happens when someone browses something very private (planning a surprise engagement. looking at medical data. planning a protest)? All that data gets slurped to google and subject to a warrant or discovery or building your advertising fingerprint.
Maybe the idea is that the data is sent to AI only when you right click, but that seems like a very thin firewall that a product manager will breach in the interests of delivering "predictive AI" via some kind of precomputed results.
Profit!
Its wild that they even put this out as a demo. It should have been picked apart in the internal meeting. There is no way I'd ever show my product taking 5s to change a 1 to a 2 in a piece of text that the user was already hovering or taking 10s to drag and drop a line of text from one box to another. Even the image of finding a route between two images could be done quick if images were auto OCR'd which is a setting on most image viewers.
Now you get to hear every person in the office do that around you.
Like, good tech, but do googlers live in the real world? Do they genuinely like the idea of an open office full of people talking to their computers? Do they all live alone without human contact?
I'm sure Don Hopkins can tell you a long annotated tale about the NeWS pizza ordering app that displayed a real-time dynamically-updated rotating pie on the screen as you filled out your order.
I like text selection exactly how it is. I want precise controls.
It's fine for a touch interface like a phone, but on a computer I expect precision. As much as I can get.
the agent occasionally spots your real problem like an experienced engineer
I'm imagining a webpage with a link - instead of opening a new link to quickly google something or opening three new tabs based on hyperlinks, i can point at a paragraph or line and ask it to tell me about it.
Maybe I can point at a song on Spotify and have it find me the youtube video, or vice versa (of course this is assuming a tool like this wouldn't stay locked into one ecosystem.. which it will).
Point is that the concept of talking to the computer with mouse as pointer is pretty cool and i guess a step closer to that whole sci-fi "look at this part of the screen and do something"
Maybe something with the file system? Like hover over any file and instead of seeing a snapshot or some details/metadata you could get a quick 1-2 line sentence on what it is. I suppose you may want to have that saved somewhere as well to cache it... but I'm def not an expert.
This would solve the always difficult issue of finding that one document! I still have trouble with document search on both mac & windows OS sometimes.
Anyway, I built a prototype on this idea, but instead of relying only on hover, I press Option to select a node in a custom AST-ish semantic layer I designed around a minimalist UI grammar, and Option + up/down arrows to move to parent/child node. This way, I have have an accurate pointer to the element I want to talk about, plus a minimal context window (parent component, state, a few navigation related queries).
What I learned from using it, though, is that the killer use case isn't necessarily the flashy "talk to this UI element" interaction shown in the Google demos. I do use it that way too; I have `Option + Shift + click` to copy a selector to the clipboard, so I can give an LLM connected to the live medium a precise reference to the element I want to discuss.
But the place where it has been most useful day to day is much simpler: source navigation. Point at the thing in the UI, jump to the code that is responsible for it. The difficult part is jumping to the code you care about (the code for UI or for the semantic element?), but in my system that distinction turned out to be usually obvious, which is what makes the interaction useful.
Imagine trying to convince someone in the 90s that that's a step forward.
Also featured in the Starfire vision video from 1992: https://youtu.be/jhe1DFY-SsQ?t=286
And the paper https://dam-prod.media.mit.edu/uuid/8e6d934b-6c6f-48e4-b0a1-...
https://en.wikipedia.org/wiki/MacPlaymate
>The game features a panic button that when clicked on will cover the computer screen with a fake spreadsheet. The player can also choose to print out Maxie's current pose as a pinup.
https://archive.org/details/mac_MacPlaymate
Geraldo interviews Chuck Farnham about getting sued by Playboy:
https://news.ycombinator.com/item?id=42571845
https://www.upi.com/Archives/1989/02/09/Playboy-sues-over-se...
Of course learning proper cad software is probably the right thing here, but having Claude write python scripts which generate HTML files which reference three.js to provide a 3d view has gotten me surprisingly far. If something could take my pointer click and reverse whatever coordinate transforms are between the source code and my screen such that the model sees my click in terms of the same coordinate system it's writing python in, well that would be pretty slick.
Until then I've just had it list every surface in a legend, each colored differently, so I can say "three inches down from the top of pole six, and rotate it so the hoop part of the bolt faces northwest."
If we manage that, my plans for a pure XML based shell might not be too futuristic '<in><ls/></in><out><tree><file date="CDATA[...]">' ...
I prefer keyboard operation myself, but I can see how this could become useful in the future, for certain use cases.
What would be useful as well then is if you could bind such a repeatedly-used AI command to button/menu item/keyboard shortcut in a way that it can still be used with pointing “this” and “that”.
On a less serious note, the audience for this is people who want to optimize for what seems like the least amount of effort.
More $$$ for the PM who launched the product.
Interesting but not “reimagining” anything.
I think the real story here is how vibe coding now enables flashy demo sites like this to be built for a concept that hasn’t yet earned it.
(Not going to happen)
Maybe you can share a scenario for that one? I can’t figure a scenario where all of this needs to be true. It seems like a recipe for accidents.
In my experience, any combination of computers + speech + danish has, so far without exception been terrible. Last time I tested ChatGPT, it couldn't understand me at all. I spoke both in my local dialect and as close to Rigsdansk [π] as I could manage. Unusable performance, and in any case I should be able to talk normally, or there's no point. It was about a year ago - it may have improved but I doubt it. I'm completely done trying to talk to machines.
Pre-emptive kamelåså: https://www.youtube.com/watch?v=s-mOy8VUEBk
https://www.media.mit.edu/publications/put-that-there-voice-...
The Siri voice transcription is pretty awful compared to what I've experienced with ChatGPT though and it's weird going back almost to the pre-LLM world where you have to give such clear sort of computer-coded voice commands.
IMO, that’s the anti social part. Why is a phone butting in your relation with your kid? I only have a few pictures taken when I was a kid (and most were for some grand occasion). And I’m happy that was the case because of so many cringeworthy things I did.
realistically what would be best is flexible inputs. voice input is often neutered by being only voice, or requiring a click activation and click to stop recording, is part of the trouble.
Getting a "magic" input is not as hard as it seems, if you reduce its input space as not needing to compete with keyboards and voice. a workflow could involve an assistant making suggestions and then the magic input needs only be a yes/no which could be head nod / shake, eye tracking & blinking, foot pedals, hand gestures
honestly though i think the real quality of life improvement is going to come from OS's enabling multiple focus windows to be active simultaneously with multiple input devices. like what i really want is a keyboard that can act as 10+ virtual keyboards with a way to change which one is in use, and then based on which one is in use the inputs go directly to an app that is in focus for that input. let my game controller stay inputting into the game while i type something, or toggle my voice input to talk to an AI and not transmit to discord while im doing that. or i just get two keywords and two mice and splitscreen games with m&k input with another person next to me to two instances of the game running (one on each of two monitors), or the mic recognizes my partners voice as a separate input from my voice and managed independently
there is so much that could be done with interface beyond layout and menus even regardless of AI, but AI related tools could help with mapping different voices into different virtual inputs, or recognizing keywords to do the same
It's like a hidden curse of LLMs -- they're so good at parsing intended meaning from non-grammatically-correct language that we don't have to be very good at clear communication.
Eventually all LLMs will be controlled by humans uttering terse gutteral grunts. We will all become neanderthals, with machines that deliver our every whim.
I recommend the youtube channel @afadingthought to see what people come up with (like v=283-z29TXeM).
Except for the large majority of people who read, type, and click way faster than they can talk. Especially for visual things it’s way faster to drag a rectangle than to describe what you want.
A lot of us also aren’t linear verbal thinkers. It would take minutes to hours to verbalize concepts we can grasp visually/schematically in seconds.
Great book on the topic: https://www.goodreads.com/book/show/60149558-visual-thinking
I usually convey the same meaning with 80wpm typing. Makes it faster to read too
Maybe I’m just slightly adhd – listening to people talk drives my crazy. Get to the point! Much easier if they type it out
Also, I doubt DeepMind is designing for existing programmers and savvy computer users. They are thinking about the other billions of people in the world. Speech is the skill people will already have, not typing.
Neither typing speed nor dictation speed is a true bottleneck, but editing speech seems like it'd be harder than editing text.
Though there may be some hybrid approach that can work well.
You don't have to think about the design of your app. You just say what you want and the AI makes it appear. If you don't like something, you tell the AI to change it. You iterate live until you get the final result you want.
This is what writing docs has become for me. I have the agent make a draft, then tell it which sections to rewrite, combine, etc. I tell it the ideas I forgot to include. I manually make certain word choice changes. The question is how do you extend this flow to non-pure-text scenarios. For most people, just talking about what you see if probably the easiest.
I hadn’t realized until just now how accurate that is for me as well. Thank you.
People have so many verbal tics and filler words too. Anthropic’s Dario says “you know” after every third word, for example.
Or they meander around unrelated/unimportant details.