Comment by aurareturn
4 hours ago
One thing seems for certain is that OpenAI models hold a distinct lead in academics over Anthropic and Google models.
For those in academics, is OpenAI the vendor of choice?
4 hours ago
One thing seems for certain is that OpenAI models hold a distinct lead in academics over Anthropic and Google models.
For those in academics, is OpenAI the vendor of choice?
OpenAI specifically targeted Academia a lot and gave out a lot of free/unlimited usage to top academics and universities/researchers.
They also offer grants you can apply for as a researcher. I'm sure other labs may have this too but I believe OpenAI was first to this.
Hasn't AlphaFold been used to make real discoveries for a few years now?
I think he's talking about reasoning models.
I think the mathematicians on X are all using GPT 5.5 Pro
From my limited testing, Gemini can dig out hard to find information given you detail your prompt enough.
Given that Google is the "web indexing company", finding hard to find things is natural for their models, and this is the only way I need these models for.
If I can't find it for a week digging the internet, I give it a colossal prompt, and it digs out what I'm looking for.
This is my experience too. Gemini and Gemini deep research are awesome. Claude's deep research is pretty bad really relative to ChatGPT or Gemini. Overall, I still love Claude the best but it is not what I would want to use if I wanted to really dig into deep research. The export to google docs in Gemini deep research is tough to beat too. I haven't used Gemini since January but have probably years of material from saved deep research in google docs. Almost an overwhelming amount of information when I dive into what I saved.
Gemini seems better trained for learning and I think Google has made a more deliberate effort to optimize for pedagoical best practices. (E.g. tutoring, formative feedback, cognitive load optimization)
As far as academic research is concerned (e.g. this threads topic), I can't say.
Agreed I usually use Gemini for explaining concepts and ChatGPT for getting things done on research projects.
Yes, I meant academic research.
Gemini is like someone with short-term memory loss; after the first response, it forgets everything. That being said, I have checked multiple model and gemini can sometime give accurate answer.
Gemini is a series with a lot of individual models.
What you are describing doesn't match my experience at all with Gemini 3 or 3.1, especially the pro version.
A simpler explanation is that more people are using ChatGPT
OpenAI models seem to have been trained on a lot of auto-generated theorem proving data; GPT 5.5 is really good at writing Lean.