Skip to content
TheChatArchiveThe Chat Archive
DeepSeekdeepseek-r1-0528· DeepSeekHugging Face LMArena Human Preference 140K

whats an interesting way to input text without using a typical keyboard setup. i am not…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with o3-2025-04-16’s answer.
  1. User

    whats an interesting way to input text without using a typical keyboard setup. i am not looking for basic ones that everyone might know like stegonography or speech recognition. as a guide, a good answer would be something like dasher.

  2. DeepSeek Assistant

    Here are several innovative text input methods that go beyond standard keyboards, speech recognition, or steganography, with Dasher-like creativity as a benchmark. Each leverages unique interaction paradigms:

    1. Neural Handwriting Decoding (BCI)

    • How it works: Users imagine writing by hand while wearing a non-invasive EEG cap or implanted electrodes. AI decodes brain signals associated with handwriting motions into text.
    • Why it's interesting: Achieves ~90 characters/minute (Stanford, 2021) by translating subconscious motor imagery into text—no physical movement required. Ideal for paralysis.
    • Example: Stanford's BrainGate BCI system.

    2. Tongue-Driven Keyboard (Tongue Tracking)

    • How it works: A magnetic tongue stud or inertial sensor tracks tongue movements in 3D space. Users "flick" their tongue to navigate a radial keyboard layout or gesture alphabet.
    • Why it's interesting: Enables hands-free input with high precision. Used in assistive tech (e.g., Tongue Drive System). Speed: ~12 wpm after training.
    • Demo: Similar to EyeWrite but for oral gestures.

    3. Breath Pattern Typing (Sip-and-Puff Morse)

    • How it works: Users inhale/exhale into a tube ("sip" = dot, "puff" = dash) to input Morse code. AI translates sequences into text with auto-completion.
    • Why it's interesting: Minimalist, accessible for quadriplegic users. Combines ancient encoding (Morse) with modern prediction.
    • Tools: Morse Code to Text translators with sip-puff hardware.

    4. Facial Expression Typing (EMG/Computer Vision)

    • How it works: Electrodes detect subtle facial muscle contractions (e.g., eyebrow raise = "select," cheek twitch = "next"). Alternatively, cameras track micro-expressions.
    • Why it's interesting: Ultra-discreet—usable in AR glasses or for users with limb disabilities. Faceteq explored EMG for gaming/text input.
    • Speed: ~5-10 wpm with prediction.

    5. Haptic Glove Sign Language Translation

    • How it works: Sensor-lined gloves capture hand/finger positions during sign language. AI maps gestures to text in real-time.
    • Why it's interesting: Bridges communication gaps for Deaf users. Projects like SignAloud or Google's MediaPipe enable this.
    • Innovation: Extends beyond text input to full language translation.

    6. Pulse-Triggered Scanning Keyboard

    • How it works: Users wear a PPG (photoplethysmography) sensor. Intentional heart rate spikes (e.g., via breath-holding) trigger selections in a scanning keyboard layout.
    • Why it's interesting: Uses involuntary biofeedback for control. Experimental but viable for locked-in syndrome.
    • Research: BCIs using hemodynamic responses.

    7. 3D Gesture Typing (LiDAR/ToF Sensors)

    • How it works: Users "draw" words in mid-air using finger motions tracked by depth sensors (e.g., Kinect). AI recognizes gestures as letters (like invisible graffiti).
    • Why it's interesting: Enables text input in AR/VR without surfaces. Project Soli demonstrated millimeter-wave radar for this.
    • Speed: ~20 wpm with practice.

    8. Chemical/Molecular Typing (Biohacking)

    • How it works: DNA strands or chemical solutions encode text. Users mix pre-defined compounds to form "letters" read by sequencers (e.g., Harvard’s Molecular Morse Code).
    • Why it's interesting: Sci-fi meets reality—stores text in molecules. Not real-time but revolutionary for data storage.

    Key Themes:

    • Assistive Focus: Many methods (BCI, tongue, breath) empower users with motor disabilities.
    • Unconventional Sensors: Leverage EMG, LiDAR, PPG, or biochemical interfaces instead of keys.
    • Hybrid AI: Combines gesture/prediction for efficiency, like Dasher.

    For a Dasher-like experience, explore Neural Handwriting Decoding (BCI) or Tongue Tracking—both prioritize speed and novelty while being actively researched. If you're prototyping, start with open-source BCI frameworks like OpenBCI or facial EMG tools like EMGduino.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.