People keep telling me that #ChatGPT is amazing for proofreading text and improving scientific writing.
I just gave #GPT4 a section of a grant proposal and it made 11 suggestions, none of which were worth keeping (often adding or removing a comma, or repeating a preposition in a list).
More interestedly, a number of its suggestions were identical to my originals.
Today, both Mailchimp and Hootsuite (very popular tools which I happen to use for @thunderbird newsletters and social media) invited me to use #AI to compose Facebook posts, tweets, and emails, respectively.
I won't be using AI tools for writing. But you should know that it's starting to get exponentially more difficult knowing if the content you're absorbing is from a human or from something like #ChatGPT.
@caseynewton built a custom AI copy editor in what sounds like just a few minutes using OpenAI's beta "GPT builder" tool.
You might be surprised to learn that's not what worries me most.
What keeps me up at night is the firehose of all kinds of automated content we're going to face, including large, fully automated influence campaigns like he describes near the end of this article.
Hello all les #mastonautes. Je cherche des profs d'#universite et du #superieur (belges de préférence) qui accepteraient de discuter de leur stratégie pédagogique face à #chatgpt... ou de leur désarroi 🤣🤣🤣 pour un article pour le prochain @curseurs. #enseignement#pedagogie. Merci d'avance pour le repouet!
#ChatGPT and its ilk are effectively a large-scale DDoS attack on not only specific tasks (such as teaching, software maintenance and fiction editing) but on human creative inspiration, sense of self and ability to make meaning of the universe.
notice how they are
creating work for us rather than saving us work
automating inherently energizing and rewarding tasks such as making art rather than tiring and tedious tasks such as filing taxes
flooding social networks with generated content that makes it harder to connect with real humans and build community
distracting us with questions like “what if this machine is sentient?” which keep us from noticing and acting to prevent the suffering of actual humans
that thing where all search engines have become basically unusable because they’re stuffed full of AI garbage? i fear that’s about to happen to everything.
i don’t believe this is necessarily intentional, but no machine that learns under capitalism can imagine another world.
A publisher claims to have staff writers to write a book for me, and because I am an unknown author, they will also market the book I did not write for me. All I have to do is loan them my name and suggest a topic.
Kenyan content moderators who reviewed incredibly graphic material to train ChatGPT are asking the Kenyan government to investigate the exploitative conditions they were subject to, which included “psychological trauma, low pay and abrupt dismissal.”
Two suggestions for interacting with ChapGPT and other LLMs:
Don’t speak to it as if it were a person. It will answer as though it is, which can lead to confusion.
Whatever question you ask, mentally preface it with the phrase, “Generate some text that sounds like a plausible answer to the following, even if you have to make stuff up.” (Because that’s what it’s going to do.)
Noteworthy: a new study comparing #ChatGPT vs doctors for responding to patient queries demonstrated clearcut superiority of #AI for improved quality and empathy of responses @JAMAInternalMed
Ever since playing with ChatGPT, I've become sensitized to the way false rationality sounds ... there's a particular vibe to what is basically coherent nonsense. And now I've started to notice when people do it too. I get this crawly ChatGPT feeling when somebody is obviously making up an "authoritative" answer to a question they know nothing about. #AI#chatgpt#psychology
#Alexa got connected to #ChatGPT, and started responding to people with disinformation. #VoiceAssistants are #interfaces, and they have the power to amplify what data sources they are interfaced with.
Who controls what information is presented by a voice assistant? Who is liable for misinformation?
Thirty minutes ago in my webinar, I told people they have to watch out for incorrect AI-generated stories on sites they used to trust, and now here's another example from MSN:
How Microsoft is making a mess of the news after replacing staff with AI
I had an unsettling experience a few days back where I was booping along, writing some code, asking ChatGPT 4.0 some questions, when I got the follow message: “You’ve reached the current usage cap for GPT-4, please try again after 4:15 pm.” I clicked on the “Learn More” link and basically got a message saying “we actually can’t afford to give you unlimited access to ChatGPT 4.0 at the price you are paying for your membership ($20/mo), would you like to pay more???”
https://www.peterkrupa.lol/wp-content/uploads/2024/01/image-4.pngIt dawned on me that OpenAI is trying to speedrun enshitification. The classic enshitification model is as follows: 1) hook users on your product to the point that it is a utility they cannot live without, 2) slowly choke off features and raise prices because they are captured, 3) profit. I say it’s a speedrun because OpenAI hasn’t quite accomplished (1) and (2). I am not hooked on its product, and it is not slowly choking off features and raising prices– rather, it appears set to do that right away.
While I like having a coding assistant, I do not want to depend on an outside service charging a subscription to provide me with one, so I immediately cancelled my subscription. Bye, bitch.
But then I got to thinking: people are running LLMs locally now. Why not try that? So I procured an Nvidia RTX 3060 with 12gb of VRAM (from what I understand, the entry-level hardware you need to run AI-type stuff) and plopped it into my Ubuntu machine running on a Ryzen 5 5600 and 48gb of RAM. I figured from poking around on Reddit that running an LLM locally was doable but eccentric and would take some fiddling.
Reader, it did not.
I installed Ollama and had codellama running locally within minutes.
https://www.peterkrupa.lol/wp-content/uploads/2024/01/image-6.pngIt was honestly a little shocking. It was very fast, and with Ollama, I was able to try out a number of different models. There are a few clear downsides. First, I don’t think these “quantized” (I think??) local models are as good as ChatGPT 3.5, which makes sense because they are quite a bit smaller and running on weaker hardware. There have been a couple of moments where the model just obviously misunderstands my query.
But codellama gave me a pretty useful critique of this section of code:
https://www.peterkrupa.lol/wp-content/uploads/2024/01/image-7.png… which is really what I need from a coding assistant at this point. I later asked it to add some basic error handling for my “with” statement and it did a good job. I will also be doing more research on context managers to see how I can add one.
Another downside is that the console is not a great UI, so I’m hoping I can find a solution for that. The open-source, locally-run LLM scene is heaving with activity right now, and I’ve seen a number of people indicate they are working on a GUI for Ollama, so I’m sure we’ll have one soon.
Anyway, this experience has taught me that an important thing to watch now is that anyone can run an LLM locally on a newer Mac or by spending a few hundred bucks on a GPU. While OpenAI and Google brawl over the future of AI, in the present, you can use Llama 2.0 or Mistral now, tuned in any number of ways, to do basically anything you want. Coding assistant? Short story generator? Fake therapist? AI girlfriend? Malware? Revenge porn??? The activity around open-source LLMs is chaotic and fascinating and I think it will be the main AI story of 2024. As more and more normies get access to this technology with guardrails removed, things are going to get spicy.
Kevin Roose so desperately wants to live in the futures tech companies are selling that he’ll eagerly do their PR for them and buy into whatever illusions of intelligence they put in front of him so he can trick himself into believing they’ll actually be realized this time.
“The field of artificial intelligence has been running through a boom-and-bust cycle since its early days. Now, as the field is in yet another boom, many proponents of the technology seem to have forgotten the failures of the past – and the reasons for them.”
“Most people are not aware of the resource usage underlying ChatGPT,” Ren said. “If you’re not aware of the resource usage, then there’s no way that we can help conserve the resources.”
In July 2022, the month before OpenAI says it completed its training of GPT-4, Microsoft pumped in about 11.5 million gallons of water to its cluster of Iowa data centers, according to the West Des Moines Water Works. That amounted to about 6% of all the water used in the district, which also supplies drinking water to the city’s residents.