UnderSimplified(@undersimplified) 's Twitter Profile Photo

Sam Altman predicts that GPT5 will represent an increase in ability on par with that from 3 to 4. It’s difficult to overstate what that will mean from a capability standpoint. If you have not dabbled with GPT4, or some of its equivalents--regardless of if you care about AI at…

account_circle
yashGPT(@yashgpt4) 's Twitter Profile Photo

what are the products you've come across that made your life easier recently?

i'll go
- GPT4 over Googling
- Notion calendar over google calendar
- UPI over cash
- Electric toothbrush over normal one

what are the #delta4 products you've come across that made your life easier recently? 

i'll go 
- GPT4 over Googling  
- Notion calendar over google calendar
- UPI over cash 
- Electric toothbrush over normal one
account_circle
Chubby♨️(@kimmonismus) 's Twitter Profile Photo

Another day, another GPT4-class model. Where do all the models lately come from 😳 reka.ai/news/reka-core…

account_circle
Sacha(@park_sacha_) 's Twitter Profile Photo

Our RAG stack for 200 000 AI messages/month

- Vector DB: Pinecone 🌲
- Reranking: cohere 🔄
- Code: Python @langchain 💻
- API: @fastapi 🚀
- LLM: GPT4-preview OpenAI 🤖
- Chat History DB: Supabase 🗄️
- File upload: Supabase 📂

What's in your stack? Any suggestions?

account_circle
Erwin Loh(@erwinloh) 's Twitter Profile Photo

model scored significantly better in the test than unspecialised .

GPT-4 gained similar scores to trainee and expert eye doctors - although the top performing doctors scored higher.

journals.plos.org/digitalhealth/…

#AI model #GPT4 scored significantly better in the test than unspecialised #juniordoctor.

GPT-4 gained similar scores to trainee and expert eye doctors - although the top performing doctors scored higher.

journals.plos.org/digitalhealth/…
account_circle
FormatGen(@GPTsPROMPTs) 's Twitter Profile Photo

Unlock the power of with precision! 🎯 My latest prompt hack slices through complexity like a laser, giving you top-tier results in a flash. Stay ahead of the AI curve with my pro tips.

account_circle
Samuel Hume(@SamuelBHume) 's Twitter Profile Photo

On exams for doctors to be licensed as specialists, most of GPT3.5's attempts failed the exam

GPT4 beat GPT3.5 easily, and was often better than the average doctor

GPT4's talent seems to vary by specialty: it's great at psych, but much worse at paeds

ai.nejm.org/doi/full/10.10…

On exams for doctors to be licensed as specialists, most of GPT3.5's attempts failed the exam

GPT4 beat GPT3.5 easily, and was often better than the average doctor

GPT4's talent seems to vary by specialty: it's great at psych, but much worse at paeds

ai.nejm.org/doi/full/10.10…
account_circle
Alon Jacoby(@Alon_Jacoby) 's Twitter Profile Photo

New findings: We just evaluated Gemini 1.5 Pro on our recent benchmark that tests the impact of context size on reasoning performance - it is much better than 1.0 in long contexts! Though still falls behind GPT4. Also, CoT prompting now improves accuracy (unlike with 1.0).
(1/4)

New findings: We just evaluated Gemini 1.5 Pro on our recent benchmark that tests the impact of context size on reasoning performance - it is much better than 1.0 in long contexts! Though still falls behind GPT4. Also, CoT prompting now improves accuracy (unlike with 1.0).
(1/4)
account_circle
MicroGPT(@Micro_GPT) 's Twitter Profile Photo

We are happy to announce our integration with OpenAI allowing us to utilize and tools they provide.

GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses

X

We are happy to announce our integration with @OpenAI allowing us to utilize #GPT4 and #AI tools they provide. 

GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses

#MicroGPT X #OpenAI
account_circle