Introduction

Billy Markus, the inventive mind behind the original meme-laden cryptocurrency Dogecoin, has taken to his X platform account "Shibetoshi Nakamoto" to share his less-than-enthused feelings about the latest version of ChatGPT. Known for his witty comments on everything from politics to tech, Markus's remarks this time were drenched in disappointment. OpenAI recently introduced ChatGPT-4o, which, in spite of big promises, seems to have left Markus wishing for something more. What's the beef with this supposedly revolutionary AI upgrade?

OpenAI rolls out ChatGPT-4o upgrade

OpenAI recently unveiled the GPT-4o model, marking a new chapter in the story of artificial intelligence. Here, the “o” stands for “omni,” indicating the bot's newfound ability to handle audio, video, and text inputs simultaneously in real time. This enhancement is touted as a move towards more natural human-computer interactions, with GPT-4o having the ability to blend images, text, video, and audio to produce innovative and unique outputs. An added feature allows GPT-4o to engage vocally, responding human-like to user queries. As if that weren't enough, devices equipped with GPT-4o can communicate amongst themselves using their built-in speakers. However, it seems not everyone is thrilled with these developments.

Features of GPT-4o

So, OpenAI recently dropped their swanky new upgrade, GPT-4o. And for those scratching their heads, the "o" stands for "omni." Essentially, this means the AI is taking a victory lap and proclaiming it can handle audio, video, and text all at once, in real-time no less! Imagine if Google Assistant decided to have a conversation with your Netflix, and Siri chimed in with a snide remark. It's like a tech party where your devices get to chat amongst themselves using in-built speakers. How "Black Mirror" can that get? This update is supposed to be a big leap toward more "natural" interactions between us humans and our feisty silicon overlords. Oh, and did I mention that this thing can reshuffle different content types to whip up something entirely new? Yeah, it's like giving a blender to a hyperactive toddler who's also a Picasso wannabe.

Markus' Experiment with ChatGPT-4o

In a twist that's more amusing than a Doge meme, Billy Markus, co-creator of Dogecoin, decided to take GPT-4o for a test spin. Spoiler alert: He wasn’t exactly handing out gold stars. Markus, under his Twitter alias "Shibetoshi Nakamoto," thought he'd challenge the AI to concoct a funny meme, then make it progressively funnier. What he got was, well, let's just say less LOL and more "meh." He tweeted images of the underwhelming memes, accompanied by the caption, “It’s uh, yeah.” Clearly, the AI's sense of humor didn’t jive with his. Imagine asking a comedian to tell a joke, then asking them to top each one until you’re rolling on the floor... except you’re not. Markus' feedback was a series of escalating prompts like “make it 10x funnier” and “the funniest thing anyone has ever seen forever,” concluding with an anti-climactic "Anyway, I didn’t laugh.” Ouch, GPT-4o, that’s gotta sting!

Disappointment with GPT-4o output

Alright, folks, gather 'round! Let's talk about how ChatGPT-4o, the latest and supposedly greatest AI by OpenAI, might've just missed the mark, at least according to Billy Markus, one of the Dogecoin founders. Markus, who is known for his sharp wit and active presence on X (formerly known as Twitter), took the new AI for a spin with...underwhelming results. Yep, even the mighty can falter!

OpenAI recently unveiled this new version of GPT-4, with 'o' standing for "omni." Sounds fancy, right? The whole idea is that this model can handle text, audio, and video in real time, aiming to create a more natural human-computer interaction. Imagine a world where your tech not only talks back but also listens and watches. Ultra-progressive, but, as it turns out, not so funny.

Marcus decided to put GPT-4o to the test by asking it to do what should have been in its wheelhouse—creating a funny meme. He then upped the ante, repeatedly asking the bot to make it funnier each time. The result? A thread of images and comments from Billy that can be summarized with a resigned “it’s uh, yeah.” Side note, is there a more gentle way to convey disappointment?

After commanding the bot to progressively make the meme funnier, he reached the point where he asked it to create "the funniest thing anyone has ever seen forever." The punchline? It didn’t even get a giggle from Markus. In his final, nonchalant dismissal, he ended the thread with “Anyway, I didn’t laugh.” Not earth-shattering feedback for a model hailed as groundbreaking.

The whole episode is a gentle (and somewhat humorously ironic) reminder that while AI advancements are indeed impressive, humor—a very human and often elusive quality—might still be beyond the robotic grasp of GPT-4o. So, if you’re looking for belly laughs, best stick to the human comedians for now.

Conclusion

To wrap it up, it's clear that even cutting-edge tech like ChatGPT-4o has its limits, at least when it comes to tickling the funny bone of a Dogecoin founder. While the integration of audio, video, and text seems like a step towards a sci-fi future, for now, humor remains a distinctly human territory. Maybe next version, OpenAI?