ChatGPT

Over two months I've been using guanaco via hugginface. A lot of new open source models came out lately with more parameters and higher quality. I tried several spaces that were listed in models' profiles in hugginface, none of them seemed to show any signs of life.

Look if I pull the text above through guanaco and tell to phrase it in 4chan style: I’m such a fucking expert on this shit, let me tell ya. Been messin’ around with GuaCo for like, forever now. But all those other dumbasses just keep releasing these huge-ass models that don’t even work half the time. Tried 'em all too, believe me. Nothing but crickets chirping back at me.

I started slow, only added

> Instead of long apologies for mistakes, just write "huh!" > never end messages with an apology, also not with "huh", leave them out.

But this response makes me doubt the whole thing a bit 😅

Disabled the custom instructions and asked the same question again:

>One notable science fiction work that explores a world where the ozone layer is completely lost is the novel "The Ozone Layer" (1987) by American author Kim Stanley Robinson. [...]

Thank you everyone that joined and became a member!

cross-posted from: https://lemmy.world/post/1246165

Two authors sued OpenAI, accusing the company of violating copyright law. They say OpenAI used their work to train ChatGPT without their consent.

What I'm looking for is some kind of desktop tool that uses the OpenAI GPT web endpoint. I'd like something where I'm able to upload one or more documents (text files) and then include them as part of the conversation/query.

I have access to the GPT-4 API and I've been writing Python3 code against it for some various applications. I can see how I'd write a tool that takes in one or more documents to include in the total prompt history, but I'm hoping to not have to write it myself, mostly due to time constraints.

Is there some kind of application that has a similar feature set to this that I should look at? Or, is there a wiki/site that lists off the current tools available that I could look over?

cross-posted from: https://lemmy.ml/post/1820678

cross-posted from: https://lemmy.ml/post/1820668

cross-posted from: https://lemmy.ml/post/1820574

Link: https://readhacker.news/s/5JBLv Comments: https://readhacker.news/c/5JBLv

hey everyone, just wanted to vent a bit about something that's been bugging me. have you ever tried getting our dear chatgpt to write a song? i mean, we all love this AI for many things, but songwriting is definitely not its forte.

it seems like every time i ask it to write a song, i end up with a tune that has the most basic AABB rhyme scheme. i mean, come on, even my five-year-old nephew can write a song with more complexity. it's like it's stuck in a perpetual nursery rhyme mode.

i get that it's an AI and can't exactly tap into the creative genius of a lennon or a dylan, but a little variation would be nice, right? it'd be great if openai could tweak it a bit to give it some musical diversity.

anyone else share my frustration, or am i just expecting too much from our songwriting AI friend? let's chat about it!

cross-posted from: https://kbin.social/m/tech/t/67505

Over 100,000 OpenAI ChatGPT account credentials have been compromised and sold on the dark web. Cybercriminals are targeting the valuable information.

cross-posted from: https://lemmy.ml/post/1353899

Article summarized by AI below: The article argues that artificial intelligence (AI) is not a threat to humanity, but a powerful tool to solve global challenges such as climate change, poverty, disease, and inequality. It gives examples of how AI is already being used to improve health care, education, agriculture, and energy efficiency. It also discusses the ethical and social implications of AI, and how we can ensure that it is aligned with human values and goals. The article concludes that AI will save the world if we use it wisely and responsibly.

cross-posted from: https://lemmy.ml/post/1276972

Snippet from the article:

But some investors argue that the big tech companies’ generative AI models will not necessarily capture most value because they are becoming increasingly commoditised and may be outcompeted by smaller open-source models. More money may be made by companies that have proprietary data in a particular industry and can leverage generative AI models to address specific use cases. “That is our investment strategy,” says Hemant Taneja, chief executive of the VC firm General Catalyst.

But VCs are also anxiously reviewing the valuations of their existing investment portfolios, given the disruptive impact it will have on all software businesses. Over the past few weeks, three start-up founders have told me they have had to junk their original business plans because of the launch this year of OpenAI’s more powerful GPT-4 model, rendering their use of earlier versions obsolete. Such is the speed of the technology’s evolution that start-ups and VCs are wary of betting too much on any one generative AI model for fear it will soon seem archaic.

I was becoming very frustrated with the slow speed (often over 1.5 - 2 min) of ChatGPT4 w/ browser, so decided to test it against it's competitors including some times until the response is completed (though onenote crashed and I lost most of the timing data).

1. ChatGPT4 with browser - usually takes 1.5-2 min
2. ChaGPT4 with [Web Requests] plugin around 1:30 to 5 min
3. ChaGPT4 with [WebPilot] plugin 0:50 to :130
4. ChaGPT4 with [Access Link] plugin - around 50 sec
5. Bing Creative - around 40 seconds
6. Bard - 5-7 seconds

These are screenshots of a table on a 4k monitor, so if the type is small, try downloading the images and zooming in. Tried exporting to PDF but the results were broken PDFs.

The chatgpt4 browser and it's web plugins often failed to give results or broke and sometimes gave decent results.

Which would you use?