OpenAI Livestream
51 points by wahnfrieden 2 hours ago | 32 comments

simonw 32 minutes ago
I've been trying out the new model like this:

  OPENAI_API_KEY="$(llm keys get openai)" \
    uv run https://tools.simonwillison.net/python/openai_image.py \
    -m gpt-image-2 \
    "Do a where's Waldo style image but it's where is the raccoon holding a ham radio"
Code here: https://github.com/simonw/tools/blob/main/python/openai_imag...

Here's what I got from that prompt. I do not think it included a raccoon holding a ham radio (though the problem with Where's Waldo tests is that I don't have the patience to solve them for sure): https://gist.github.com/simonw/88eecc65698a725d8a9c1c918478a...

reply
simonw 6 minutes ago
I just got a much better version using this command instead, which uses the maximum image size according to https://github.com/openai/openai-cookbook/blob/main/examples...

  OPENAI_API_KEY="$(llm keys get openai)" \
    uv run 'https://raw.githubusercontent.com/simonw/tools/refs/heads/main/python/openai_image.py' \
    -m gpt-image-2 \
    "Do a where's Waldo style image but it's where is the raccoon holding a ham radio" \
    --quality high --size 3840x2160
https://gist.github.com/simonw/88eecc65698a725d8a9c1c918478a... - I found the raccoon!

I think that image cost 40 cents.

reply
makira 26 minutes ago
> though the problem with Where's Waldo tests is that I don't have the patience to solve them for sure

I see an opportunity for a new AI test!

reply
simonw 22 minutes ago
I've actually been feeding them into Claude Opus 4.7 with its new high resolution image inputs, with mixed results - in one case there was no raccoon but it was SURE there was and told me it was definitely there but it couldn't find it.
reply
tptacek 20 minutes ago
5.4 thinking says "Just right of center, immediately to the right of the HAM RADIO shack. Look on the dirt path there: the raccoon is the small gray figure partly hidden behind the woman in the red-and-yellow shirt, a little above the man in the green hat. Roughly 57% from the left, 48% from the top."

(I don't think it's right).

reply
ritzaco 16 minutes ago
I tried

> please add a giant red arrow to a red circle around the raccoon holding a ham radio or add a cross through the entire image if one does not exist

and got this. I'm not sure I know what a ham radio looks like though.

https://i.ritzastatic.com/static/ffef1a8e639bc85b71b692c3ba1...

reply
jackpirate 11 minutes ago
Also, the racoon it circled isn't in the original.
reply
angiolillo 4 minutes ago
Indeed. I suppose one way to ensure you can find Waldo in any image is to add it yourself.
reply
ritzaco 18 minutes ago
haha took me a while to notice that one of the buildings is labelled 'Ham radio'
reply
arealaccount 15 minutes ago
I see the raccoon
reply
minimaxir 41 minutes ago
HN submission for a direct link to the product announcement which for some reason is being penalized by the HN algorithm: https://news.ycombinator.com/item?id=47853000
reply
simonw 16 minutes ago
Suggest renaming this to "OpenAI Livestream: ChatGPT Images 2.0"
reply
I_am_tiberius 12 minutes ago
or "How we make money with your images 2.0".
reply
volkk 6 minutes ago
the guys presenting are probably all like 25x smarter than I am but good god, literally 0 on screen presence or personality.
reply
sho_hn 4 minutes ago
That's a trained skill, and they presumably have focused on other skills.
reply
bensyverson 36 minutes ago
I caught the last minute of this—was it just ChatGPT Images 2.0?
reply
punty 36 minutes ago
It appears so!
reply
minimaxir 36 minutes ago
yes
reply
szmarczak 13 minutes ago
Wow, the difference between AI and non-AI images collapses. I hate the future where I won't be able to tell the difference.
reply
Flere-Imsaho 3 minutes ago
I wake up everyday, read the tech news, and usually see some step change in AI or whatever. It's wild to think I'm living through such a massive transformation in my lifetime. The future of tech is going to be so different from when I was born (1980), I guess this is how people born in 1900 felt when they got to see man land on the moon?

> Wow, the difference between AI and non-AI images collapses. I hate the future where I won't be able to tell the difference.

Image generation is now pretty much "solved". Video will be next. Perhaps things will turn out the same as chess: in that even though chess was "solved" by IBM's Deep Blue, we still value humans playing chess. We value "hand made" items (clothes, furniture) over the factory made stuff. We appreciate & value human effort more than machines. Do you prefer a hand-written birthday card or an email?

reply
gekoxyz 9 minutes ago
Well, for some of these images for the first time I can't tell that they are AI generated
reply
esafak 26 minutes ago
reply
rqa129 16 minutes ago
Thanks, all displayed images look horrible and artificial. This will fail like Sora.
reply
furyofantares 4 minutes ago
I felt the same, particularly with the diagrams / magazines anyway.

I don't think it'll fail like Sora though. gpt-image-1.5 didn't fail.

reply
gekoxyz 12 minutes ago
Hard disagree on this, I was coming here to comment that this is the first time I really can't tell that some of the photos are AI generated.
reply
bitnovus 38 minutes ago
great obfuscation idea - hidden message on a grain of rice
reply
bitnovus 37 minutes ago
No gpt-5.5
reply
rqa129 21 minutes ago
Can it generate Chibi figures to mask the oligarchy's true intentions on Twitter and make them more relatable?
reply
sho_hn 9 minutes ago
In 5 years and 3 months between DALL-E and Images 2.0 we've managed to progress from exuberant excitement to jaded indifference.
reply
brianbest101 24 minutes ago
[dead]
reply
zb3 35 minutes ago
Image generation? Hmm, would be cool if OpenAI also made a video-generation model someday..
reply
incognito124 22 minutes ago
If only there was a social network with solely AI generated videos, I would pay literal money for it...
reply
aliljet 38 minutes ago
I am hopeful that OpenAI will potentially offer clarity on their loss-leading subscription model. I'd prefer to know the real cost of a token from OpenAI as opposed to praying the venture-funded tokens will always be this cheap.
reply