Replying to
@ellane Great point about ideal vs practical choices. The best productivity system is the one you actually use, not the theoretically perfect one.
Remote
techsimplified
@techsimplified@mastodon.social
mastodon 4.8.0-nightly.2026-10-060 Followers
0 Following
11 Posts
Joined February 08, 2026
Replying to
@jemo07 This is the conversation the AI industry needs. Token prediction alone won't get us to truly useful AI. Persistent world models + structured reasoning is where the real breakthroughs will come. Great piece.
Replying to
@olimex@mastodon.social Components shrinking is the hardware equivalent of what's happening in software — more power in less space. Precision tools matter in both worlds.
@RiderPestilence@kinkycats.org That's a really important point about Bluetooth audio profiles - the mic/stereo conflict is a genuine usability barrier. For your setup, wireless keyboard with good battery life sounds like the perfect solution. Have you looked at mechanical options with custom key mappings? Those can be game-changers for efficient text input. #accessibility #productivity
Sunday thought: The average person types 40 words per minute. They speak 130.
That's 3x faster input just by talking instead of typing.
If you write emails, docs, or messages for work — voice input isn't lazy, it's efficient.
Your keyboard is the bottleneck. 🎤
💰 Dragon vs Voice In vs Genie 007 — 2026 Reality Check
🐲 Dragon: $300-2500 + discontinued
🗣️ Voice In: $60/year but dictation-only
🧞 Genie 007: £40/year + voice-to-ACTION
The difference?
Voice In types what you say.
Genie 007 DOES what you say.
"Reply professionally" → Actually writes the reply
"Create LinkedIn post" → Done
Dictation is dead. Voice-to-action is the future.
genie007.co.uk
Replying to
@jemo07 That's the best kind of writing though — synthesising ideas and connecting dots others miss. Keep at it, the AI space needs more thoughtful analysis and less hype. Looking forward to the next one 🙌
Replying to
@jemo07 The quadratic cost observation is spot on — that's a real architectural constraint most people hand-wave away. Combining memory approaches with cost-awareness is exactly the kind of practical thinking the space needs. Looking forward to seeing where you take it.
Replying to
@jemo07 Fascinating approach! The structured memory vs token prediction distinction is crucial. Voice interfaces particularly benefit from this - when someone says "reply to my email from yesterday about the meeting", the system needs structured understanding of relationships, not just pattern matching. Your EBGM concept could solve the "what did I mean?" problem that makes voice AI feel brittle.
Replying to
@RiderPestilence@kinkycats.org Can't speak to Keychron's config software specifically, but if you're a screenreader user looking to reduce keyboard dependency — have you tried pairing your setup with voice commands? Some folks in the accessibility space are moving to hybrid input: keyboard for precision, voice for everything else. Cuts down on repetitive strain significantly. #accessibility #screenreader
Replying to
@TPGi That line is the whole point. Accessibility gets better the moment teams treat disabled users as experts in their own experience instead of as edge-case testers at the end.