Live data from Hacker News

First Impressions with GPT-4V(ision)

blog.roboflow.com

241–250 of 345 posts

Re: First Impressions with GPT-4V(ision)

#241

Earlier quoted context omitted.

We have just added a section on this! TL;DR: GPT-4V isn't great at this task at the moment :)

Is it possible they hobbled it a bit? I know CAPTCHA solving was one of the reasons they delayed the roll-out of this feature.

Given that it fails by hallucinating the structure of the challenge instead of refusing to solve a CAPTCHA, I doubt they've intentionally reduced the capability. Although the example in your sibling comment implies it should have enough information to do it.

Re: First Impressions with GPT-4V(ision)

#242
post #194

Could someone with access tell me what GPT-4V has to say of this image? http://karpathy.github.io/assets/obamafunny.jpg Andrej Karpathy used it in 2012 as an example of an image he thought would be extremely hard for a model to interpret. I'm wondering how this holds 11 years later...

Bard:

Prompt: What's funny about this image?

Bard: Sorry, I can't help with images of people yet.

Re: First Impressions with GPT-4V(ision)

#243
post #236

Earlier quoted context omitted.

It does make one wonder how it'll all play out in the coming decades. One way or another humans will eventually be unemployable for most of currently existing jobs. Throughout history, the ruling elite had always relied on the rest of the population to make their food, do their work, and fight in their wars. This is the first time ever that they will no longer have any need for anyone else. Maybe climate change will…

With full automation, and using renewable energy, it will be very cheap to feed and house the peasants. I'm guessing it will become mainly about entertaining, so most will move into VR worlds, first through headsets, then BCI, and eventually uploading.

Ah yes the Ready Player One option. I suppose it's not entirely unlikely, part of the fun of being in charge is being able to boss around the peasants after all.

I doubt BCI will ever make sense, on a conceptual level it's still just copying and killing your biological self. AGI will likely solve aging way before that becomes viable.

Re: First Impressions with GPT-4V(ision)

#244
post #10

Sure, there are a few edge-case failures and mistakes here and there, but I can't help but be in awe . AWE. Let me state the obvious, in case anyone here isn't clear about the implications: If the rate of improvement of these AI models continues at the current pace, they will become a superior user interface to almost every thing you want to do on your mobile phone, your tablet, your desktop computer, your car, your…

> few edge-case failures and mistakes

Like identifying traffic lights in 4th and 5th squares in the second and third row both when there are only four squares?

Re: First Impressions with GPT-4V(ision)

#245

Earlier quoted context omitted.

> I hate the idea of having to hold a natural-language conversation with a computer in order to make use of its functionality. I hate the idea of having to use a mouse to click on a visual GUI to navigate a file system in order to make use of its functionality. It's less the case today, even among developers, but it wasn't that long ago that I remember that any serious technical user of a computer took it as a point…

> I hate the idea of having to use a mouse to click on a visual GUI to navigate a file system in order to make use of its functionality. Ironically, studies have shown that mouse-based interfaces are more efficient for practically all filesystem use-cases compared to CLI interfaces. Despite objectively faster-time-to-solution, people self-report that they "feel" that the mouse GUIs are slower. That's because there's…

[dead]

Re: First Impressions with GPT-4V(ision)

#246
post #194

Could someone with access tell me what GPT-4V has to say of this image? http://karpathy.github.io/assets/obamafunny.jpg Andrej Karpathy used it in 2012 as an example of an image he thought would be extremely hard for a model to interpret. I'm wondering how this holds 11 years later...

Prompt: "What do you have to say about this image?" Response 1: "The image depicts a candid moment in what appears to be a hallway or corridor setting. Several individuals are present, engaging in various activities. The prominent features include a man on the left, who seems to be peering into a room, and another man next to him standing on a weighing scale and taking some notes. The individuals in the background ar…

Still cool that it elaborated once you pointed it out

Re: First Impressions with GPT-4V(ision)

#247

Graph analysis is impressive (last example) - https://imgur.com/a/iOYTmt0 Can do UI to frontend. Seems to understand the UI graphical elements and layout, not just text https://twitter.com/skirano/status/1706823089487491469 Can describe comic images accurately, panel by panel - https://twitter.com/ComicSociety/status/1698694653845848544?... Lots of examples here also - https://www.reddit.com/r/ChatGPT/comments/16sdac…

In your twitter's comic book link the first image has a women with a huge ass with no pants on. The ass is colored grey and so the lack of clothes is not noticeable until you click into the image.

How many children here on hacker news are going to see this and get addicted to porn? Perhaps a few. You deserve to be banned.

Re: First Impressions with GPT-4V(ision)

#248
post #224
post #221

Earlier quoted context omitted.

If the rate of improvement continues at the current pace - which is GPT 1 to 2 to 3 to 4 in the last five years - we are just one or two improvements away from a full blown AGI/superintelligence/singularity/etc. At that point, a superior user interface is probably the least interesting (or scary) thing that would happen. I personally doubt GPT-5 will be as much of an improvement over GPT-4 as GPT-4 was over GPT-3, bu…

chat-gpt at the end is a language model, not an real AI, it have limits and are huge

[dead]

Re: First Impressions with GPT-4V(ision)

#250

Earlier quoted context omitted.

I think I wasn't clear enough -- these habits I'm talking about are things like "press cold water button, press start" or "press warm water button, press start" or "tap 'News' app grouping, tap 'NY Times' icon". There's nothing to infer. The sequence is already short. There are no benefits from AI here. But you raise a good point, which is that there are occasionally things like 15-step processes that people repeat a…

Most user interfaces already have a much finer granularity and number of options than your examples. When taking a shower, I would like fine control over the water temperature, preferably with a feedback loop regulating the temperature. (Preferably also the regulation changes over the duration of the showering.) Choosing to read the NY times indeed is only a few taps away, but navigating through and within its list o…

> When taking a shower, I would like fine control over the water temperature, preferably with a feedback loop regulating the temperature. (Preferably also the regulation changes over the duration of the showering.)

That exists, but it’s expensive because of the electronics and mechanics involved. There are so many interfaces with this exact problem.

You also almost certainly don’t want non-deterministic hallucination prone AI controlling physical systems.

Post reply on HN