Hacker News Reader

Comment by energy123 | Hacker News Reader

  > There's no evidence of a technological moat or a competitive advantage in any of these companies.

I disagree based on personal experience. OpenAI is a step above in usefulness. Codex and GPT 5.2 Pro have no peers right now. I'm happy to pay them $200/month.

I don't use my Google Pro subscription much. Gemini 3.0 Pro spends 1/10th of the time thinking compared to GPT 5.2 Thinking and outputs a worse answer or ignores my prompt. Similar story with Deepseek.

The public benchmarks tell a different story which is where I believe the sentiment online comes from, but I am going to trust my experience, because my experience can't be benchmaxxed.

View on HN

I still find it so fascinating how experiences with these models are so varied.

I find codex & 5.2 Pro next to useless and nothing holds a candle to Opus 4.5 in terms of utility or quality.

There's probably something in how varied human brains and thought processes are. You and I likely think through problems in some fundamentally different way that leads to us favouring different models that more closely align with ourselves.

No one seems to ever talk about that though and instead we get these black and white statements about how our personally preferred model is the only obvious choice and company XYZ is clearly superior to all the competition.

There is always a comment like this in these threads. It’s just 50-50 whether it’s Claude or OpenAI.

We never hear what the actual questions are. I reckon it's Claude being great at coding in general and GPT being good at niche cases. "Spikey intelligence"

I’m not saying that no company will ever have an advantage. But with the pace of advances slowing, even if others are 6-12 months behind OpenAI, the conclusion is the same.

Personally I find GPT 5.2 to be nearly useless for my use case (which is not coding).

Can you give specific examples? I'm super interested to see where it fails.

For me OpenAI is the worst of all. Claude code and Gemini deep research is much much more better in terms of quality while ChatGPT hallucinating and saying “sorry you’re right”.

codex is sooo slow but it is good at planning, opus is good at coding but not at good at seeing the big picture

I use both and ChatGPT will absolutely glaze me. I will intentionally say some BS and ChatGPT will say “you’re so right.” It will hilariously try to make me feel good.

But Gemini will put me in my place. Sometimes I ask my question to Gemini because I don’t trust ChatGPT’s affirmations.

Truthfully I just use both.

I told ChatGPT via my settings that I often make mistakes and to call out my assumptions. So now it

1. Glazes me 2. Lists a variety of assumptions (some can be useful / interesting)

Answers the question

At least this way I don't spend a day pursuing an idea the wrong way because ChatGPT never pointed out something obvious.

Care to share the system prompt?

Reply on news.ycombinator.com