this post was submitted on 17 Apr 2024
257 points (89.5% liked)
Technology
60009 readers
3806 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Nobody should be surprised by this, and I don't see how it's "fake" at all.
Systems like this are extremely error prone. There's no way you can get an acceptable level of accuracy without extensive human review. Doesn't mean there's no AI — there is. It's just the AI is merely to help those humans do their job.
You're completely missing the point.
It's portrayed as all AI to get investors, but in reality most of the "human help" end up having to do like 90% of the work.
But the company runs around telling potential investors lies to get money
To be fair, the bulk of AI advancements aren't visible to anyone but the people closest to the code who were doing machine learning / data analysis anyway.
All of the "magic" of it that's somehow swindling billions out of venture capitalists because it's going to replace so many people is made up hype garbage. Yeah, it can write the same paragraph on any subject you choose. Hooray. Also, that's not really helpful unless not giving a shit is part of the communications process.
It's replacing human scammers, I guess. There's that.
It's not tho...
That's why Amazon cancels their "just walk out to checkout" program.
They kept saying it was AI, but then eventually admitted they were paying overseas workers to do it via webcam because AI couldn't do it.
They kept that program going for years before they gave up on it
Not those scammers, the Nigerian Princes.
The most popular venues for GPT assisted confidence schemes right now is a tie between discord and twitter. Twitter allows for the utilization of short form response bots in DM and posts, discord has an ongoing AI generated art scam where a robot begs you to comission them so they can make rent. Both have extremely easy to identify playbook/flowchart type responses, and key messages they will always send to push the scam along among their generated chatter. It's not quite nigerian prince, and it's only getting more prominent as neither site has a handle on the current con and thus aren't doing anything to curb it
I worked in the object recognition and computer vision industry for almost a decade. That stuff works. Really well, actually.
But this checkout thing from Amazon always struck me as odd. It's the same issue as these "take a photo of your fridge and the system will tell you what you can cook". It doesn't work well because items can be hidden in the back.
The biggest challenge in computer vision is occlusion, followed by resolution (in the context of surveillance cameras, you're lucky to get 200x200 for smaller objects). They would have had a really hard, if not impossible, time getting clear shots of everything.
My gut instinct tells me that they had intended to build a huge training set over time using this real-world setup and hope that the sheer amount of training data could help overcome at least some of the issues with occlusion.