I haven’t found ChatGPT’s composition to be all that compelling or well-done. I mean, it’s impressive that the technology can do it at all, but does anyone have an example of anything like a 3k word story that is well-written and fun to read with a unique voice? Not saying this could never happen (with future models), but every time I’ve prompted a story it’s pretty weak and obviously AI-generated. What am I missing? #ChatGPT#WritingCommunity#fiction#AI#MachineLearning#LLM
Before you install the new #ChatGPT app on iOS, ask yourself: what's the worst thing the maker of the world's greatest bullshit machine could do with this kind of information, and why are they requesting it?
For the first time today I heard the phrase "this sounds like it was written by #ChatGPT" to describe a blog post that was verbose, boring and full of platitudes.
this story that "someone is using AI to spread disinformation about Changpeng Zhao of Binance" is going around crypto media, in reports mostly based on this Fortune Crypto article.
what is the claim here, that someone somehow intentionally seeded ChatGPT with false info? that congresspeople are asking ChatGPT if CZ is a communist and then sending formal inquiries? that someone is sending congresspeople ChatGPT printouts with false info, and they're taking them seriously?
#Chatgpt fails at the most basic stuff, like converting milligrams to kilograms. Who on Earth would want to use it for anything meaningful? #artificialintelligence
A few of my many thoughts on this hot mess of a GPT story, relayed by @kendraserra:
These events really give the lie to the putative imminent replacement of professionals by AI. I’m quite certain some VC out there has money sunk into “LLMs will replace lawyers!” I know they’re trying with coders, therapists, tech support…and…just look at this frigging mess.
It's about ChatGPT, lawyers, and a very angry judge. The details are juicy, and the implications are huge. Let's dig in. 🍿
⚖️ 3 characters are important: Roberto Mata (plaintiff), Steven Schwartz (attorney), Peter LoDuca (attorney on record).
Mata sues Avianca airlines through Schwartz at state court. Avianca transfers the case to Manhattan's federal court. This is when things get interesting. 🕵️♂️
Sorry, ich kann es nicht anders sagen, aber unsere Kinder werden mit einer KI-Shitshow aufwachsen. Und es liegt an jedem Einzelnen von uns, diese KI-Shitshow nicht so schlimm werden zu lassen, wie ich sie mir aktuell vorstelle. 🥶
Crawler für Training von GPT ausschließen: OpenAI stellt Code für robots.txt vor
Für das Training von KI-Chatbots wie ChatGPT wird das Internet durchpflügt. Wer Inhalte davon ausschließen will, kann das nun über Code in der robots.txt tun.
I do not recall an open letter with thousands of prominent signatories, a hastily-assembled White House Task Force and a big Senate hearing for automated driving systems.
You know... #SafetyCritical systems that are masquerading as #AI that have killed people and have the capacity to readily cause immediate injury and death.
We have an unregulated Wild West out there on that.
Everyone seems to be caught up in the AI Utopia hopium, when they should be paying attention to the error rate of ChatGPT.
OpenAI's ChatGPT can actually make workers perform worse, a new study found
and before anyone cries "but they used GPT-3. GPT-4 is clearly better. Not fair!". They tested GPT-4.
To test this, researchers randomly assigned 758 consultants at BCG across one of three groups: one with no access to AI; one with access to ChatGPT powered by GPT-4; and one with access to ChatGPT, as well as instructional videos and documents on prompt engineering strategies.
The #ChatGPT resumes have started to appear in my clinics and mentorship sessions, and boy.... howdy, kiddos. Always remember to use a screwdriver to turn screws and a hammer to hammer nails. AI is a tool that you should apply carefully in the right places.
It's good at formatting bullet points and being a thesaurus, but not good at writing a solid and coherent narrative about a person in a way that will sell the candidate. I have to do a lot of corrections on these to account for interviewer attention span and flow. Be very careful.
I have a preprint out estimating how many scholarly papers are written using chatGPT etc? I estimate upwards of 60k articles (>1% of global output) published in 2023. https://arxiv.org/abs/2403.16887
How can we identify this? Simple: there are certain words that LLMs love, and they suddenly start showing up a lot last year. Twice as many papers call something "intricate", big rises for "commendable" and "meticulous".
Next week I'll be starting a pretty ambitious project—50 Days of LIT Prompts. Every weekday for 10 weeks, I'll be sharing prompt patterns along with my thoughts and readings relating to Large Language Models like those behind #ChatGPT. Follow the link below, and this thread, for updates: https://sadlynothavocdinosaur.com/posts/50-days-of-lit-prompts/
AI-Generated Data Can Poison Future AI Models - Scientific American (archive.is)
As AI-generated content fills the Internet, it’s corrupting the training data for models to come. What happens when AI eats itself?