AI image generators, and their errors, tell you very little about LLMs & their capabilities. When you ask ChatGPT or Bard to produce an image, the LLM is just writing prompts (often inexpertly) and those prompts are then processed separately by the relatively dumb image makers.
@emollick
-
Gemini Claude ChatGPT Bing interfaces tools comparison
By
–
A few more comparisons:
Interfaces: Gemini feels the most polished, Claude is extremely stripped down and feels the most clean, and Bing & ChatGPT are… what they are. Tools: GPTs are a big advantage for GPT-4. Gemini has interesting connections to Google. Claude is alone. -
Comparing Claude 3, Gemini Advanced, and GPT-4 strengths
By
–
As for differences:
1) "Personality" varies, you will likely find one more appealing than the others
2) Each has strengths and weaknesses. Claude 3 is a good write, Gemini Advanced a good explainer, GPT-4 is all around solid.
3) Features such as context windows, multimodal, etc. -
Claude 3: GPT-4 Class Model Performance Review
By
–
Yes, I was writing about Claude 3. I have had advanced access for a couple of days. It is a GPT-4 class model, both in benchmarks and in use. Better in some ways, worse in others.
-
Three GPT-4 Class LLMs Share Surprising Similarities
By
–
We now have three GPT-4 class LLMs. Their overlap is a bit surprising:
1) All the LLMs prompt in very similar ways. The minor variations will not matter to most folks
2) All three have roughly similar rates of hallucination, about similar things
3) They are all a little pedantic -
Cultural Fit Hiring Perpetuates Homogeneity in Elite Firms
By
–
Hiring for cultural fit just means hiring yourself. Interviewers judge fit by comparing candidates to themselves. The paper shows candidates at elite law, banking & consulting firms are "good fits" if they have similar hobbies, styles & college experiences as their interviewers
-
Claude Outperforms in Multimodal Graph Analysis Study
By
–
Multimodal & graphs: Looking at a set of four graphs from a study of law students & AI, Claude does by far the best, Gemini 1.5 does well, and GPT-4 has trouble with visual details, leading to some hallucinations. However, when asked to provide the means for the graph, all fail.
-
Creative Commons AI Prompts for Educators and Students
By
–
Here are our prompts for educators and students, released under Creative Commons license. They are built around the pedagogical approaches we discuss in our papers, and have worked well in class. I hope others experiment with them, improve them, & share. https://
moreusefulthings.com -
Why GPT-4 Remains Competitive Against Claude 3 Gemini
By
–
How if GPT-4 still a top model? Claude 3 and Gemini Advanced do exceed it in some areas, but it is still pretty close Does it mean:
1) GPT-4 is as good as it gets (I suspect not)
2) OpenAI has The Secret
3) Everyone else stopped training when they beat GPT-4, better models soon -
Claude 3 vs GPT-4: Benchmark Performance Comparison Analysis
By
–
Useful to note that Claude 3 does not definitely surpass GPT-4 on benchmarks; which have improved over time since GPT-4’s release (Claude 3 compared itself only to the initial benchmarks). This matches my personal experience, its a GPT-4 class model, with some special strengths.