• TOKEN EFFORT
  • Posts
  • If AI was Marvel, then ChatGPT would be Iron Man

If AI was Marvel, then ChatGPT would be Iron Man

AI is coming for our jobs and everything we hold dear. Not fun.

Subscribe at www.tokeneffort.io

AI is coming for our jobs and everything we hold dear. Not fun.

But I thought it’d be a good idea to trivialise things (while still learning) and turn 8 of AI’s biggest products into Marvel characters! Fun!

AI has been having a moment since ChatGPT launched in November 2022, and it’s only snowballed since. Getting a handle on everyone who does what, where, when and how is too much for anyone who isn’t a large nerd.

But it’s important to at least get the basics down… Even if it’s just to look some of these companies in the eye as their tech vaporizes massive chunks of what people used to do for a living. *smiling face with single tear emoji*

Note, this is just the tip of the AI-ceberg (apologies). There are hundreds of companies riding this wave of AI, and this is just a silly way to remember a handful of the biggest, and a bit of their “vibe”.

ChatGPT (part of OpenAI) is IRON MAN

The original and the best. ChatGPT has all the gadgets, all the publicity, and genuinely the best product. A flawed hero with their share of bad press, but that comes with the territory when you’re the main character. Personally, I’m all in on ChatGPT.

Microsoft Copilot (part of Microsoft) is WAR MACHINE

Just a shit version of Iron Man isn’t it. Copilot runs off OpenAI’s LLM (same stuff powering ChatGPT) but in plain clothes and can’t do all the stuff ChatGPT can. Copilot is what your IT department is telling you to use instead of the good ones.

Gemini (part of Google) is THANOS

Gemini is… inevitable. Backed by the hugeness of Google’s Search, YouTube, Gmail and Android. It’s setting up to take over the world, but not quite yet. Launching after ChatGPT, plus some bugs mean they’re still playing catch up. As in the movies, Iron Man and friends could kill Thanos but equally likely in our universe though, Thanos kills them.

Meta AI (part of Meta/Facebook) is THE HULK

Meta AI has been Bruce Banner-ing in the background for a bit, but with “LLaMA” it’s large language model (Meta AI’s “brains”) now shoved into everything from Facebook to WhatsApp to Ray-Bans, Meta AI could soon go full Hulk smash on us. Remember Meta is second only to Google in terms of global scale.

Claude (part of Anthropic) is SPIDERMAN

Small and nimble like Spidey, but keep an eye on Claude/Anthropic as with big money backing it’s one of the big 3 LLM’s (alongside OpenAI and Google’s products). A bit holier than thou, Anthropic was started by ex-OpenAI (Chat GPT) employees who are all about that “with-great-power-comes-great-responsibility life”. They created Anthropic to do more for “AI Safety” 

Perplexity (independent) is VISION

Big old nerd-burger just like Vision. Perplexity’s “thing” is that it blends real-time web search with LLM. This means it gives you sources for its very thorough answers to your questions. And that means it cuts down on the “hallucinations” or incorrect info/straight-up lies that a ChatGPT might give you. Note, Perplexity hasn’t built its own LLM, it piggybacks off others including Anthropic, Google and OpenAI.

Grok (part of Elon Musk's xAI) is DEADPOOL

Weird. Inappropriate. Definitely capable, but not shy of a bit of piss-taking. Grok to Deadpool was the easiest comparison to draw, and unsurprising since it’s part of Elon’s mad mad world. Plugged into X (formerly Twitter), Grok gives you hyper up-to-date info but is limited by the fact there’s lots of questionable content on X. Grok isn’t bad, but it’s not good either.

DeepSeek (independent) is BLACK PANTHER

Just like Black Panther, DeepSeek arrived with bang from a “secretive far away land” (China) with levels of tech The West was not expecting, causing a huge fuss by simply showing up to the party. Super advanced and does it using less resources than competitors - its also open source meaning it will be in loads of other stuff before long. 

This newsletter will be about making AI easier to digest, so we’ve started pretty light. Let me know what you think, I want to hear your thoughts and suggestions.

If you want minimal bullshit AI news and opinion, subscribe to Token Effort.