The man who killed Google Search?

honestly, 99% of the time I don't need search. I want AI. I don't want to have to use a weird syntax to 'talk to a search engine'. sometimes I don't even know what the word is that I am searching for. I want something I can just ... talk to.

I use to use search every day, now I use it about once a month.

As somebody who uses "weird syntaxes" to create applications every day, I like having the option to use a specific language that offers the ability to more precisely describe the parameters of my search.

Do you find Bing search to fit your needs or do you use something else? I honestly get tired of having to type so much to get it to find what I actually want. Most often I do prefer to just use my acquired Google-fu of speed reading results.

Indeed, AI is immensely useful! I use it every day too.

However, it's been my experience that finding original works, perhaps that I can cite as a source, is somewhat difficult when the computer might confabulate both the content and the citations.

When LLMs get (much) better at doing math, law and medicine, I'll be much more likely to use them for those things.

I would like it if they would actually respect the weird syntax consistently

@drowsspa

I don't use bing search, I use chatgpt and claude.

Here are some examples: after pasting hundreds of log lines of output from a failed build request, "why did this build fail?"

After pasting my last 3 workouts, "I am wondering if I am not putting enough muscle on my body / torso. Is this the case? if so, suggest me an exercise that utilizes body weight, dumbbells, or weighted exercise ball"

It suggested dumbbell pull over, so I asked "What weight should I start with for the dumbbell pull over?"

"say I want to go to the club and seem like I know what I am doing, how many dances should I know?"

"say I have a pandas series of numpy.ndarray, and I have an numpy.ndarray. I want to find the cosine distance between the numpy.ndarray and each of the items in the series"

"I made a notebook for non data scientists to follow and use, so I want to add lots of comments and mark down documentation." (paste notebook code) and it adds comments, doc strings, etc.

most of this stuff, using search as it is, is clunky. I would have to find weird ways to word what I am searching for to find results.

Have you tried Phind? I find that it takes most of the "clunkiness" out of the process of searching by interrogating AI + searching the web at the same time.