The litigation by a patent troll ("Voice tech corporation") seems to be real. What a complete waste of resources.

The patent troll is what killed the business plan in the end because there was zero margin. The real problem seems to me that it is difficult to compete with a nicely integrated 30 usd echo dot if people rather pay it with their personal data (this actually includes me). Mycroft also technically was falling short even for its potential target audience. They had a hard time even without that troll.

If I would do this I would rather repurpose (and potentially resell) off the shelf mobile phones with a custom stand and rather get the software right first.

I think this is a sad way to go. Hardware is super expensive to develop and the key IP here is in the software. There's probably some fancy microphone tech, but ultimately you could (should?) run Mycroft on a Raspberry Pi with an edge accelerator device.

Technically I think we're getting close to a good offline system. Home Assistant solves the practical aspect of controlling systems and triggered routines - it really works well, at least on par with vendor software. Speech recognition models are usable and getting better. The interesting "glue" is parsing the commands into something reasonable and I think open source LLMs are likely to be the answer. See Open Assistant [0] for example. It doesn't solve the knowledge base issue (e.g. people seem to use Alexa either as a very fancy egg timer, or for querying facts), but it would probably allow for very natural interaction with the device and the model could just translate a query into a command that triggers an action.

My impression is that Amazon's approach (I have no knowledge of what goes on inside) has been to build fairly complicated subsystems to handle very specific queries. Each "Skill" has to be separately developed and if you ask something outside a fairly rigid grammar, the device has no idea.

[0] https://github.com/LAION-AI/Open-Assistant

worth reading both sides: https://www.eff.org/document/voice-tech-corp-v-mycroft-ai-co...

Defendant should not have so engaged Plaintiff.

That being said, shouldn't it be trivial to find prior art; prior software or articles already published on the internet that does what USPTO allowed Plaintiff to patent (2017) after searching only its own database?

Please understand how frustrating this obvious, on a mobile PC, prior-art-invalidateable patent is:

   def on_voice_input(audio):
      text = speech_to_text(audio)
      tasks = parse_text(text) #
      for task in tasks:
          exec(task) # run locally, or
          exec("ssh user@remote_host -- $task") # run remotely

Isn't that what Hal does in "2001: A Space Odyssey" (1968)?

Pretty sure there were voice command apps for Palm, Symbian, Blackberry, WinMo, etc; years before the date on the cited patents.

(And furthermore, such has certainly been possible on desktop computers (PCs) for many years prior to 2017.) E.g. Dragon Naturally Speaking (now Nuance) has been able run to custom commands `on_voice_input()` on a PC since the early 2000s at least IIRC)

Given voice commands prior art, to then run a remote command - over SSH, for example - is an obvious next step and thus not a patentable invention to this reasonable person.

(Industry has developed open specifications for running commands on remote computers: Instead of keying SSH over IP for all devices in a LAN/WLAN, Thread (2014) and Matter (2019) transmit task messages with just AES over IEEE 802.15.4 protocols over WiFi. Thread is designed to replace existing, older protocols which lacked security features. Thread_(network_protocol)#IoT_protocols_landscape https://en.wikipedia.org/wiki/Thread_(network_protocol)#IoT_...

Thus there is also prior art for running remote commands on remote devices (whether they be PCs, Macs, Linuxes, or mobile devices)

If you can run a desktop operating system on a mobile device, is mobile device a relevant distinction?

(In my opinion, on a PC patents are vexatious.)