[FRIAM] hacking neural nets via compiler

Prof David West profwest at fastmail.fm
Wed Oct 12 21:24:17 EDT 2022


Yeah,  all I was saying is it is interesting. I share your antipathy to the various "noises" about the techTopic of the hour. I really enjoy all the inconsistencies: blockchain is so "safe" alongside the multiple scams and thefts (I was in a bookstore today and skimmed the title of a book about some woman who "stole" billions in crypto and got away with it—not a book I would even read the flyleaf).

Everything is a risk. Any attempt to hype that risk is, to cynical ol' me, is just someone trying to amass power and wealth.

davew


On Wed, Oct 12, 2022, at 4:04 PM, glen wrote:
> It's kinda weird. Way more serious vulnerabilities exist in literally every other supply chain we rely on ... like undermanned ports ... or the 3 big credit agencies ... yet if ML/AI is involved, a perverse, relatively difficult to pull off attack is elevated from the depths of dorkhood to something we should spend more time on.
> 
> I guess it's similar to the hyenas screaming about crypto. Both the advocates and their critics are sooooo LOUD. I can barely think.
> 
> Don't take this post the wrong way, though. It is a cool technique.
> 
> 
> On October 12, 2022 2:38:28 PM PDT, Prof David West <profwest at fastmail.fm> wrote:
>> interesting article
>> 
>> davew
> -. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .
> FRIAM Applied Complexity Group listserv
> Fridays 9a-12p Friday St. Johns Cafe   /   Thursdays 9a-12p Zoom https://bit.ly/virtualfriam
> to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> FRIAM-COMIC http://friam-comic.blogspot.com/
> archives:  5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/
>   1/2003 thru 6/2021  http://friam.383.s1.nabble.com/
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20221012/d1ed73a0/attachment.html>


More information about the Friam mailing list