janriemer, to random

😯Wait, what!? is not growing exponentially, but has reached it's plateau!?

How can that be!? This is such a surprise! Oh my!😱

/s

OpenAI’s CEO Says the Age of Giant AI Models Is Already Over

https://www.wired.com/story/openai-ceo-sam-altman-the-age-of-giant-ai-models-is-already-over/

Wyndix,

@janriemer Wow!!!! It's not like 99% of the fediverse has known this since the beggining

alcinnz,
@alcinnz@floss.social avatar

@janriemer But laypeople were so insistant that there were smart people working in AI, that the technology would keep improving, and the current flaws would be addressed!

/sarcasm

Sigh, they don't seem to comprehend how much innovation would be required to overcome the current flaws...

georgeeyong, to ChatGPT
@georgeeyong@mastodon.social avatar

Very powerful linguistic assessment. I bet #ChatGPT fails this.
Source: dunno, got it from green text bubble checkmark app thingy.

maegul,
@maegul@hachyderm.io avatar

@georgeeyong

I'm surprised at how much I struggled with this. You think you're going to nail it but then it gets you.

judeswae, to accessibility
@judeswae@toot.thoughtworks.com avatar

A new volume of ’ opinionated guide to software is now available. Check it out. https://thght.works/40G2ENC

Also, find out if the team used at any time during the writing of this report. ;)

teachpaperless, to random

I haven’t tried this out yet, but this looks like a show stopper for educators everywhere!

Auto Classmate | Helping Educators Navigate AI and Education https://autoclassmate.io/

natharari,

@teachpaperless Just forwarded that to a great friend of mine who is a professor. We were just talking about AI in education last Sunday.

alexbfree, to random
@alexbfree@mastodon.social avatar

My tip for :

Everyone should train yourself to imagine that every query you type into a chatbot is prefixed by

"Give me something that LOOKS LIKE an answer to the question: "

And then treat the response accordingly.

Might answer your question, might be complete bollocks.

alexbfree,
@alexbfree@mastodon.social avatar

What we currently call AIs are basically Expert Pretending Machines. Like Jacob Rees Mogg. Might look and sound convincing at times, but also highly prone to talking total twaddle.

video/mp4

joe, to random
@joe@toot.works avatar

This is kind of interesting. You can use https://haveibeenencoded.com/ to see who LLMs think you are.

https://haveibeenencoded.com/

derekb_wi,

@joe So, you're a professional golfer? I got NBA free agent 😀

joe,
@joe@toot.works avatar
feynman, to opensource

@msprout keep me in touch with the models you're working with. I'm also trying to stay ahead of the curve. #Vicuna is perfect on Apple Silicon M2P, it's almost as fast as #GPT3. I saw also #huggingface made an Open Source alternative to #ChatGPT but it's overloaded.

https://huggingface.co/chat/

#FOSS (but not on purpose thanks Meta!)

samsteiner, to random German
@samsteiner@swiss.social avatar

Used #AI to create a #Spotify #playlist

1️⃣ asked #ChatGPT what a good image prompt would be for a catchy cover image for a Indie Rock playlist
2️⃣ created the image with that prompt
3️⃣ used Spotify algorithm (search) to find a song list
4️⃣ asked ChatGPT for a list of top Indie Rock songs
5️⃣ used randomize.org do shuffle the list (not AI)

Very basic usage.

AI could do more, like analysing the moods, the tempi, the BPMs, the danceability and the keys (via Spotify API) to fine-tune a playlist...

samsteiner,
@samsteiner@swiss.social avatar
sebmeineck, to random German
@sebmeineck@mastodon.social avatar

speichert neuerdings nach eigenen Angaben jegliche Konversation mit für einen Monat und schaut bei Bedarf rein. Und das ist ein Datenschutzplus. Zuvor habe OpenAI mit den Konversationen das Modell auch weitertrainiert.

https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt

liztai, to random
@liztai@hachyderm.io avatar

The problem with #AI is not the AI itself but the humans who use it.

I think that's the gist of the problem with #ChatGPT and #Midjourney and everything else.

Used responsibly and with care, it may be a force for good.

Unfortunately many will use it as a way to get a quick buck at the expense of others.

absolutspacegrl,
@absolutspacegrl@mastodon.social avatar

@liztai we’ve learned nothing, rinse wash repeat.

liztai,
@liztai@hachyderm.io avatar

@absolutspacegrl Greed trumps all. Sigh

Mastokarl, to random
@Mastokarl@mastodon.social avatar

Tomorrow I will destroy #chatgpt by asking it how to get a teenage kid to keep his room tidy. I fully expect it to first run into a frenzied endless loop, trying to solve this devious impossible problem, and then to sink into a catatonic freeze, not answering any further questions.

So, time to ask your final questions.

😜

Quokka,

@Mastokarl I turn off Wi-Fi for his devices 💻 📵

itnewsbot, to random
@itnewsbot@schleuss.online avatar

ChatGPT now allows disabling chat history, declining training, and exporting data - Enlarge (credit: OpenAI / Stable Diffusion)

On Tuesday, OpenAI... - https://arstechnica.com/?p=1934271

Tekchip, to random
@Tekchip@mastodon.social avatar
liztai, to random
@liztai@hachyderm.io avatar
alienghic,

@misc @liztai

I'm sorry, but I don't want to produce offensive content.

misc,
@misc@mastodon.social avatar

@alienghic @liztai You’re a good Bing. ☺️

garry, to random

Today’s AI News & Comment

Stability AI launches StableLM, an open source ChatGPT alternative
https://arstechnica.com/information-technology/2023/04/stable-diffusion-for-language-stability-launches-open-source-ai-chatbot/

Human values, as well as AI, must be at the core of the future of work
https://www.theguardian.com/global-development/2023/apr/25/human-values-as-well-as-ai-must-be-at-the-core-of-the-future-of-work-chatgtp

Opera One is a new AI-powered browser that aims to beat Chrome and Edge
https://www.techradar.com/news/opera-one-is-a-new-ai-powered-browser-that-aims-to-beat-chrome-and-edge

“GPT” may be trademarked soon if OpenAI has its way
https://techcrunch.com/2023/04/24/gpt-may-be-trademarked-soon-if-openai-has-its-way/

Greywing’s new SeaGPT solves email overwhelm for maritime crew managers
https://techcrunch.com/2023/04/24/greywing-seagpt/

How to improve your ChatGPT prompts and responses
https://www.techradar.com/how-to/how-to-improve-your-chatgpt-prompts-and-responses

U.S. policing AI at companies to make sure it doesn't violate civil rights
https://www.reuters.com/technology/us-policing-ai-companies-make-sure-it-doesnt-violate-civil-rights-2023-04-25/

OpenAI Offers New Privacy Options for ChatGPT
https://www.bloomberg.com/news/articles/2023-04-25/openai-offers-new-privacy-options-for-chatgpt?leadSource=uverify%20wall

A developer exploited an API flaw to provide free access to GPT-4
https://techcrunch.com/2023/04/25/a-developer-exploited-an-api-flaw-to-provide-free-access-to-gpt-4/

AI-Generated Amazon Reviews Are Beginning to Spread
https://gizmodo.com/amazon-ai-reviews-chatgpt-chatbot-gpt4-1850374323

Apple is reportedly developing an AI-powered health coaching service
https://techcrunch.com/2023/04/25/apple-is-reportedly-developing-an-ai-powered-health-coaching-service/

The latest viral Drake AI hit is just a repurposed SoundCloud rap
https://www.theverge.com/2023/4/25/23696155/viral-drake-ai-repurposed-soundcloud-rap

News app Artifact can now summarize stories using AI, including in fun styles
https://techcrunch.com/2023/04/25/news-app-artifact-can-now-summarize-stories-using-ai-including-in-fun-styles/

GOP Launches the Age of AI-Generated Attack Ads
https://gizmodo.com/youtube-biden-ai-political-ad-gop-stable-diffusion-1850373150

Spotify CEO says AI progress is both “really cool and scary,” may pose risk to creative industry
https://techcrunch.com/2023/04/25/spotify-ceo-says-ai-progress-is-both-really-cool-and-scary-may-pose-risk-to-creative-industry/

Hugging Face releases its own version of ChatGPT
https://techcrunch.com/2023/04/25/hugging-face-releases-its-own-version-of-chatgpt/

Auto-GPT will supercharge ChatGPT and change how we use it – in very scary ways
https://www.techradar.com/news/auto-gpt-will-supercharge-chatgpt-and-change-how-we-use-it-in-very-scary-ways

Yelp Adds Video Reviews and AI Features to Its App
https://gizmodo.com/yelp-reviews-short-form-video-1850372980

‘As an AI language model’: the phrase that shows how AI is polluting the web
https://www.theverge.com/2023/4/25/23697218/ai-generated-spam-fake-user-reviews-as-an-ai-language-model

Nvidia releases a toolkit to make text-generating AI ‘safer’
https://techcrunch.com/2023/04/25/nvidia-releases-a-toolkit-to-make-text-generating-ai-safer/

AI assistants come to Alibaba and a16z-backed Cider
https://techcrunch.com/2023/04/25/ai-assistants-come-to-alibaba-and-a16z-backed-cider/

Snapchat's users seem to really hate AI
https://www.imore.com/apps/social-apps/snapchats-users-seem-to-really-hate-ai

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • kavyap
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • osvaldo12
  • khanakhh
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • everett
  • ngwrru68w68
  • Durango
  • anitta
  • InstantRegret
  • GTA5RPClips
  • cubers
  • ethstaker
  • normalnudes
  • tacticalgear
  • cisconetworking
  • tester
  • Leos
  • modclub
  • megavids
  • provamag3
  • lostlight
  • All magazines