Axe-Fx III Firmware 23.00 Public Beta #1

Status
Not open for further replies.
Dave Merrill said:
V to g

Perhaps I'm easily impressed, but I actually was impressed by that summary and its follow-up. Pretty good job actually.
Click to expand...
It’s scary and I’m scared! 😂 It’s uncanny when it’s in its wheelhouse, but frightening when it’s asked to do something that it’s not good at, because it gives answers with such confidence that you can’t tell the difference. I found it’s good at linguistic problems, but can’t even list the amps in the AFX, it just made up weird ones. Might be better now but I still worry that people will ask it to do serious things (like legal briefs, or a letter to parents after a school tragedy, not even joking), or something medical, and not be able to tell when it’s way off base.
 
I used Chat GPT and another one (can't remember the name of it) when putting together some legal papers about a month ago. I needed case law relevant to my pleadings and did a search to the effect of "cite case law with quotes on issue of fractal users obsession with fw updates" and it would give me 5 pieces (if available) of useful info. I then vetted that info and 95% of the time it was exactly what I needed. I did need to include "with quotes" to help track down the specifics in said case law, but this saved me 10's of hours of research!
 
<OT>

In my fitful sleep last night I had strange 'creative nightmare' in which I was frantically attempting to come up with a fail-safe design for proof of digital provenance/watermarking for pics/videos using encryption and hashing etc. I am no expert in any of that, but I kept seeing holes in the design. I was freaking out, like I couldn't whack all the moles to make it work without some vulnerability.

I think my dream was triggered by the confusion about the hospital bombing and how to "prove" what actually happened. I'm seriously starting to worry that very soon we won't be able to trust any pics/videos unless they have some trustable proof of origin. To mock up a satellite video of an errant or nonerrant missile strike will be cake soon and then no "authority" (eg govts, media orgs) will be trusted... not to mention exponential rampup of fake news pollution & conspiracies.

No politics here, just concern about AI further confusing and dividing humanity and even what facts are real.

I know people are working on solutions (at least in academia) but I haven't heard many alarm bells going off in govt or in the MSM.

</OT>
 
Status
Not open for further replies.
Back
Top Bottom