[-] besselj@lemmy.ca 41 points 4 weeks ago* (last edited 4 weeks ago)

Every new internet-connected feature they add seems to also increase attack surface, so this is a weird dilemma. Still better to keep things up to date in the short term. In the long term, I've made the choice to switch to an OS that respects user preferences.

Edit: I've also heard that the Windows 11 IoT enterprise edition doesnt come with all this bloat but still gets regular security updates

[-] besselj@lemmy.ca 42 points 1 month ago

Looks like the AI-first strategy is really working out for them

[-] besselj@lemmy.ca 40 points 1 month ago

My guess is that they dont want anybody to use the word 'Microsoft' in their account name to impersonate the company.

[-] besselj@lemmy.ca 40 points 4 months ago

Copy that, we're clear on opsec

[-] besselj@lemmy.ca 40 points 4 months ago

Big AI companies pretty much exclusively sell LLMs that output unreliable data, so idk how much of a worry it is anymore.

[-] besselj@lemmy.ca 40 points 5 months ago

Yup. Reading the Wikipedia article about Travis Kalanick, it seems like he just became a billionaire by coming from a privileged background and getting lucky.

[-] besselj@lemmy.ca 42 points 5 months ago

Y'all remember that #DeathTo Black Mirror episode?

[-] besselj@lemmy.ca 42 points 5 months ago

Please bro, trust me bro, just a few more billion. AGI is just around the corner bro

[-] besselj@lemmy.ca 41 points 5 months ago

I still don't understand what Anthropic is trying to achieve with all of these stunts showing that their LLMs go off the rails so easily. Is it for gullible investors? Why would a consumer want to give them money for something so unreliable?

[-] besselj@lemmy.ca 43 points 5 months ago* (last edited 5 months ago)

It's like the trump admin has never heard of the Streisand effect. Hopefully the developer doesn't get harassed or arrested and the service doesn't get taken down for a while.

People should be wary about not giving any info that might get tied back to them, just in case it's a honeypot https://infosec.exchange/@molytov/114780966893879907

I would much prefer to see a service like this being developed and hosted outside of US jurisdiction, for obvious reasons.

[-] besselj@lemmy.ca 43 points 7 months ago

So how close can you get your eyeballs to the sensor? Even if IR isn't in the visible spectrum, that doesn't mean it can't damage your eyes at high power levels. If anything, its more dangerous because you won't notice it.

[-] besselj@lemmy.ca 40 points 7 months ago

It'll be interesting to find out if this research got IRB approval prior to the study

view more: ‹ prev next ›

besselj

joined 8 months ago