Skip to content
By Kevin O'Keefe, LexBlog CEO & Founder

ChatGPT-4 a Big Step Up over GPT-3 In Both the ChatGPT and the Bing Atmospheres

Abstract arrows concept

ChatGPT-4 is a big step up over ChatGPT-3 in the both the ChatGPT and Bing atmospheres.

I have been using ChatGPT-3 regularly for a couple months. I run it on my Chrome browser and upgraded to ChatGPT Plus for twenty dollars a month.

With ChatGPT Plus, I received ready access, anytime, and the ability to make more queries in the day. Plus now gives me early access to ChatGPT-4.

On Saturday I received early access to the new Bing, Microsoft’s search engine.

Bing, running with ChatGPT-4, is a big step up from search. Better interface, slick conversational approach and a prompting of what other questions I may want to ask.

The prompt coming from knowing who I am – I’ll identify the who by situation and role – and the discussion that has ensued.

With 4, the speed is incredible. You receive an answer in seconds and you are replying a couple seconds later. Probably faster than texting with someone.

The GPT interface does not prompt you for questions you may want to ask but the stripped to a bare bones interface has me hanging with it.

Maybe that’s because Bing is out of site, out of mind, living over on the Edge browser, versus the Chrome browse, which I have open all day on my iPad, the only machine use.

Bing attributes the sources from which it culled its answers and even footnotes sources. GPT does not.

Big deal for lawyers and law firms looking to be recognized for their expertise, for bloggers looking to be recognized for their writing and for bloggers looking to see who the influencers are whom they may want to follow and reference.

Accuracy is also better on 4. I was told that in a blog post from OpenAI, GPT’s parent company, and in the coverage of yesterday’s news.

I asked questions on Bing relating to online legal publishing and such publishing being done by large law. Some of the answers referenced LexBlog and I as sources of for its answers.

When I asked what LexBlog does and for whom, the answers were shorter, more precise and more accurate than on GPT-3.

Particularly, GPT-4 was pretty accurate in its description of LexBlog’s services, whereas GPT-3 always said we offered services to write a lawyer’s blog content, something we do not.

The speed at each things are moving – and improving is pretty amazing. Can only imagine what GPT-5 will offer.