OpenAI's livestream demo of GPT-4 with Greg Brockman, showing its capabilities and limitations. Highlights include writing code, turning a hand-drawn sketch into a website, and a tax-question example.
Why Hacker News points people here: HN refers back to it as the primary source for what GPT-4 could do at launch, especially the sketch-to-website and multimodal demos.
“the TaxGPT example in the GPT-4 Developer Livestream [1] was powerful”HN comment, 2023
“the TaxGPT example in the GPT-4 Developer Livestream [1] was powerful”
TaxGPT and coding examples in the livestream
Full comment
the TaxGPT example in the GPT-4 Developer Livestream [1] was powerful. I'm hoping (soon?) for a Microsoft 365 assistant that can do the same trick on ALL your MS cloud data (docs/files/mails/notes...) Edit : the included "Business Chat" goes in the same direction (on what they call the Microsoft Graph) Also, there were interesting "coding" examples in the same livestream. So the coming GPT-4 (and 5) integration in the Power Platform [2] could be a game-changer for less technical staff. [1] this video [2] link
Cites itjonas21Mar 2023
“creating a website from a crude sketch”
Demos of coding and sketch-to-website cited as evidence of intelligence
Full comment
If passing the bar exam, writing code alongside a human and incorporating their feedback [1], or creating a website from a crude sketch [2] aren't akin to human intelligence, then what is? At some point, you have to go from saying it's just a parlor trick to considering that maybe we're just a parlor trick as well. [1] this video [2] this video
Cites itrvnxMar 2023
Timecode of Be My Eyes mention in announcement
Full comment
Maybe it's something else ? I am genuinely not sure what they mean, this is the timecode in the announcement: this video It seems to say that they use Be my eyes "to make the product better", so it could be humans in the loop during captioning, or maybe they are using volunteers data from Be my eyes as training set ? Or it's completely opposite, and that Be my eyes is using the OpenAI APIs ? My guess is that it is both ?
Cites itlonelyasacloudMar 2023
“it is shown being fed a photograph of a hand-drawn web page”
Demo turning a hand-drawn sketch into a working web page
Full comment
In one of GPT4's release demos, it is shown being fed a photograph of a hand-drawn web page. It then image analyses that, works out that it's a picture of a web page and turns it into a working web page ( this video ). So here's the thing; what in that list of non-codegen tasks is more complicated than that which GPT can already automate?
Cites itZephyrBluMay 2023
“The developer livestream was on March 14th”
Dates the GPT-4 livestream to March 14th
Full comment
I completely agree, I feel the same way as a dev. GPT-4 is not even 2 months old . The developer livestream was on March 14th: this video . The time since GPT-4 already feels something like 6 months. So far I'm perpetually feeling behind.
Cites itTenokeSep 2023
“they showed the multimodal capabilities back in the GPT4 Developer Livestream”
Shows OpenAI demoed multimodal capabilities before release
Full comment
>I wouldn't say that OpenAI "telegraphed" their intentions They did telegraph it, they showed the multimodal capabilities back in the GPT4 Developer Livestream[0] right before first releasing it. 0. this video
Cites itskilledSep 2023
“it was already revealed during the GPT-4 livestream”
Whiteboard-to-app feature was already shown in the livestream
Full comment
Not sure this is that interesting since it was already revealed during the GPT-4 livestream back[0] in March. It’s my understanding that this feature is still rolling out to everyone, so if there are some use cases beyond “simple” - I am sure we’ll hear about it in a more detailed article / blog post. Not taking anything away, just pointing out the facts. [0]: this video
Cites ittkgallyJan 2025
“Greg Brockman’s demonstration of using it to write code”
Brockman's coding demo of GPT-4
Full comment
Minor correction: GPT-4 was announced on March 14, 2023, less than two years ago. I don’t remember how much LLMs had been discussed as coding assistants before then, but it was Greg Brockman’s demonstration of using it to write code that first brought that capability to my attention: this video Advances since then have indeed been remarkable.
