Googzilla Versus OpenAI: Moving Up to Pillow Fighting

May 17, 2024

dinosaur30a_thumbThis essay is the work of a dinobaby. Unlike some folks, no smart software improved my native ineptness.

Mike Tyson is dressed in a Godzilla outfit. He looks like a short but quite capable Googzilla. He is wearing a Google hat. (I have one, but it is soiled. Bummer.) Googzilla is giving the stink eye to Sam AI-Man, who has followed health routines recommended by Huberman Lab and Anatoly, the fellow who hawks supplements after shaming gym brutes dressed as a norm core hero.


Sam AI-Man asks an important question. Googzilla seems to be baffled. But the cane underscores that he is getting old for a thunder lizard selling online advertising. Thanks, MSFT Copilot. How are the security initiatives coming along? Oh, too bad.

Now we have the first exhibition: Googzilla is taking on Sam AI-Man.

I read an analysis of this high-stakes battle in “ChatGPT 4o vs Gemini 1.5 Pro: It’s Not Even Close.” The article appeared in the delightfully named online publication “Beebom.” I am writing in Beyond Search, which is — quite frankly — a really boring name. But I am a dinobaby, and I am going to assume that Beebom has a much more tuned in owner operator.

The article illustrates a best practice in database comparison, just tweaked to provide some insights into how alike or different the Googzilla is from the AI-Man. There’s a math test. There is a follow the instructions query. There is an image test. A programming challenge. You get the idea. The article includes what a reader will need to run similar brain teasers to Googzilla and Sam AI-Man.

Who cares? Let’s get to the results.

The write up says:

It’s evidently clear that Gemini 1.5 Pro is far behind ChatGPT 4o. Even after improving the 1.5 Pro model for months while in preview, it can’t compete with the latest GPT-4o model by OpenAI. From commonsense reasoning to multimodal and coding tests, ChatGPT 4o performs intelligently and follows instructions attentively. Not to miss, OpenAI has made ChatGPT 4o free for everyone.

Welp. This statement is not going to make Googzilla happy. Anyone who plays Foosball with the beastie today will want to be alert that re-Fooses are not allowed. You lose when you what the ball out of the game.

But the sun has not set over the Googzilla computer lab. The write up opines:

The only thing going for Gemini 1.5 Pro is the massive context window with support for up to 1 million tokens. In addition, you can upload videos too which is an advantage. However, since the model is not very smart, I am not sure many would like to use it just for the larger context window.

I chuckled at the last line of the write up:

If Google has to compete with OpenAI, a substantial leap is required.

Several observations:

  1. Who knows the names of the “new” products Google rolled out?
  2. With numerous “new” products, has Google a grand vision or is it one of those high school stunts in which passengers in a luxury car jump out and run around the car shouting. Then the car drives off?
  3. Will Google’s management align its AI with its staff management methods in the context of the regulatory scrutiny?
  4. Where’s DeepMind in this somewhat confusing flood of “new” smart products?

Net net: Google is definitely showing the results of having its wizards work under Code Red’s flashing lights. More pillow fights ahead. (Can you list the “new” products announced at Google I/O? Don’t worry. Neither can I.)

Stephen E Arnold, May 17, 2024


Got something to say?

  • Archives

  • Recent Posts

  • Meta