Would be more amusing if Microsoft resurrected the "Bob" name for something AI.
Or Tay
IBM AI ('Bob') Downloads and Executes Malware
11–20 of 126 posts
Re: IBM AI ('Bob') Downloads and Executes Malware
#12These prompt injection vulnerabilities give me the heebie jeebies. LLMs feel so non deterministic that it appears to me to be really hard to guard against. Can someone with experience in the area tell me if I'm off base?
Re: IBM AI ('Bob') Downloads and Executes Malware
#13Re: IBM AI ('Bob') Downloads and Executes Malware
#14Re: IBM AI ('Bob') Downloads and Executes Malware
#15Maybe I'm paranoid, but allowing any coding agent or tool to execute commands within terminal that is not sandboxed somehow will be prone to attacks like that
Re: IBM AI ('Bob') Downloads and Executes Malware
#16Re: IBM AI ('Bob') Downloads and Executes Malware
#17These prompt injection vulnerabilities give me the heebie jeebies. LLMs feel so non deterministic that it appears to me to be really hard to guard against. Can someone with experience in the area tell me if I'm off base?
> it appears to me to be really hard to guard against
I don't want to sound glib, but one could simply not let an LLM execute arbitrary code without reviewing it first, or only let it execute code inside an isolated environment designed to run untrusted code
the idea of letting an LLM execute code it's dreamt up, with no oversight, in an environment you care about, is absolutely bananas to me
Re: IBM AI ('Bob') Downloads and Executes Malware
#18These prompt injection vulnerabilities give me the heebie jeebies. LLMs feel so non deterministic that it appears to me to be really hard to guard against. Can someone with experience in the area tell me if I'm off base?
> Can someone with experience in the area tell me if I'm off base?
Nope, not at all. Non-determinism is what most software developers write. Something to do with profitability and time or something.
Re: IBM AI ('Bob') Downloads and Executes Malware
#19I didnt know IBM was even in this game.
They got in early with Watson
Re: IBM AI ('Bob') Downloads and Executes Malware
#20These prompt injection vulnerabilities give me the heebie jeebies. LLMs feel so non deterministic that it appears to me to be really hard to guard against. Can someone with experience in the area tell me if I'm off base?
> it appears to me to be really hard to guard against I don't want to sound glib, but one could simply not let an LLM execute arbitrary code without reviewing it first, or only let it execute code inside an isolated environment designed to run untrusted code the idea of letting an LLM execute code it's dreamt up, with no oversight, in an environment you care about, is absolutely bananas to me
just wait until the exploit is so heavily obfuscated that you just review and allow it to get the project done.