To me (price aside) this tiny laptop looks like an amazing piece of gear for someone that otherwise would be looking to make a cyberdeck. I don’t see it like one of the late 2000s netbooks that were all critically underpowered, but look at it as a pre-built cyberdeck. Finally a tiny laptop with power and potentially lots of uses for the gaming controls integrated into it. This would make for an awesome SDR host (especially for those SDRs that require strong host compute), emcomm setup, maybe even some way to control a drone with the hand controllers with some middleware etc.
GPD Win Max 3
I’m saying cyberdeck very loosely, as my favorite cyberdeck builds are not the bare-bones ‘looks cool’ but actually can’t do much builds but ones like:
And by far my favorite cyberdeck build is below- its big and bulky but was built to truly be a swiss army knife of compute (even EE troubleshooting gear included- you could use his kit to reverse engineer a car’s proprietary canbus, 0-5v sensors, signal analysis and on and on.
But I see this Win Max 3 as a powerful and highly compatible start of a hardware baseline and with a nice little case of accessories be able to do a lot of things. Maybe more power hungry than desired, but then again with more and more locally hosted AI inserting into people’s workflow, maybe perfect. For example using a local LLM to filter through wireshark.
@PhaseLockedLoop I’ve been pretty dark on the forum due to work lately, but been using AI more and more for some vibe coding projects (don’t judge me).
Do you see AI making breakthroughs in the dark arts that is RF? Like compact snowflake looking antennas that perform really well at HF frequencies? Not just RX and TX as well?
I can see it augmenting the existing design knowledge and helping people build it easier and come up with some theories on how to refine those ideas but humans will have to do the grunt work of meticulously proving it and making it useful. Understanding will be critical here.
Will it create something novel winning. X doubt. All those high profile “look what AI did to prove what a scientist was about to” stuff also proves another problem. Will we understand the advance or proof?
You can’t replace knowledge. Use it as a tool. Use it to help expand the foundation of the knowledge but do the due diligence in learning whats underneath and understanding it because you’ll often find every time you do. It sometimes gets the nuances wrong
Thats the other danger. You need some base knowledge of what your doing to know if it got the nuances of understanding incorrect and sent you down a rabbit hole
So I guess my answer is Judge PLL is out rn. Try again when we have more data and concrete understood novel new proofs from ai and we understand ai well enough to guard rail it into being useful… Think.. Your problems getting it to understand you and output exactly what you want but on mass scale.. This isn’t to discourage its use. Its to encourage critical thinking of what AI can do and can’t do for us. I.e know the limitations of your tool
