F

I can't believe we're going to high pricing based subscription. This makes me think we need than ever open source models like qwen and deepseek.

On the other hand, the free options are better than ever. I've only used the free versions of ChatGPT, Claude, and ImageFX and have been impressed by how much you get without spending a dime. The only real limitation I regularly feel is not being able to upload something like a 1GB CSV file for analysis, but I guess that's a fair restriction for a free web tool.

I am impressed by the free version of ChatGPT, not so much with Claude's (claude.ai). I love Claude Pro, more so than the paid version of GPT, but damn, the free version of Claude is awful, you reach the limit extremely quickly. The free version of ChatGPT works quite well.

It’s because these companies are burning billions of dollars.

The problem is that comparable-quality open models will still require lots of hardware to run at reasonable speeds, so even then it's not cheap.

This shows how unrealistic, inaccurate and pointless coding benchmarks are.

I also include similar code interview platforms like leetcode, hackerrank and so on.

Plot twist:

1. they are trying to obfuscate deepscrape success

2. they are trying to confuse you. the benchmark margins are minimal (and meaningless)

3. they are trying to get time (with investors) releasing nothing-special-models in a predicted schedule (jan -> o3, feb -> o3-pro-max, march -> o7-ultra, and in 2026 -> OMG! we've reached singularity! (after spending $500B))

-

And at the end of the day, nothing changes for me and neither for you. enjoy your time out of this sickness ai hype. bruh!