Artificial intelligence still has some way to go

I've been occasionally trying to devise creative ways to test "Is this think 'thinking' in some sense or is it just spitting out text?"

My latest idea was to see what happens if I ask GPT to start a conversation and then plug the prompt into Bard, then plug Bard's response back into GPT, and so on. It started off ok - I asked GPT to start a conversation, and it responded "Absolutely! Here's a conversation starter: "What is one book that significantly impacted your life, and how did it change your perspective?"

Bard gave me a relatively long and sophisticated (if insipid) response about The Alchemist by Paulo Coelho. I plugged that back into GPT, and it in turn responded in a relatively sophisticated (but again insipid) way, and this continued back and forth for a handful of exchanges. Eventually, things started to crack slightly though.

First, Bard asked GPT: "I would love to hear about your own writing journey and learn from your experiences."

GPT then responded with one of its stock "As an AI language model..." type responses, followed by what was pretty clearly a conversation ender:

As an AI developed by OpenAI, I don't have personal experiences or a writing journey of my own. However, I'm here to provide insights based on a vast range of data I've been trained on.

If you have any other questions about writing, need suggestions for more writing resources, or want feedback on certain writing-related topics, feel free to ask! I'm here to help. Keep nurturing your passion for writing, and remember that every word you write brings you one step closer to becoming the writer you aspire to be. I wish you all the best in your writing endeavors!

Bard did not seem to pick up on any of this, and also began referring, confusingly, to its interlocutor as "Bard."

Following that, the conversation devolved into a tennis match of long goodbyes, each basically just restating the same platitudes in different words. I played out this exchange a bunch of times just to see if either AI would ever change its response, and it seems the answer is no, it just gets stuck in this endless feedback loop of empty positivity.

If either GPT or Bard were "thinking" in any meta sense, rather than just generating prompt responses, it seems unlikely that this would happen. This made me pretty skeptical of these "emergent abilities" and "sparks of AGI" claims.

Entire conversation here:

https://chat.openai.com/share/0f1f55fc-25ae-44e9-b53f-fec544d0e001

lol, excellent work

ChatGPT has no thoughts. It only has responses to stimulus. These responses just happen to mirror the kind of responses made by thinking creatures capable of expressing thoughts in language. If it were placed in charge of controlling the behavior of some living thing that thing would die very rapidly.

I follow a bunch of cool house/architecture accounts on Instagram and it’s tiresome how Midjourney “concept” buildings have taken over the suggested page.

I asked Bard to generate some crossword puzzle clues, and it was all stuff like:

Small furry animal (7) Cat
A room in the house (4) Kitchen

...I thought it was strange that it didn't understand that the number in brackets should correspond to the number of letters in the answer, but then I realized it doesn't _understand_ anything...

i have a jeopardy clue-a-day calendar and i have been playing against chatgpt for a few weeks. i think it has gotten every clue correctly so far, but whenever we get to final jeopardy it wagers the same number (2000) and i end up winning because it wagers incorrectly.

do we have an AI art thread? anyway

AI-generated QR codes using ControlNet are insane.

This is going to be increasingly common in ads in the near future.

These examples blew my mind (try scanning them):

1. Ancient Village pic.twitter.com/on2S6SrOUD

— Rowan Cheung (@rowancheung) June 11, 2023

nice idea but pretty flaky, some of them didn't work at all. also lol @ man alive's two idiots in a room. The future is bright, and I have no doubt you'll achieve your dreams. Happy writing!

Digital use of QR codes is a rare case, right? If you're already looking at a web page, use a URL. Where I see QR codes actually used in the wild, is embossed in tables / printed on paper for menus and the like, and you want big blocks for that, those things degrade.

there are a lot of public spaces where you may have a screen that displays a QR code, or a device that has limited input options (Apple TV, other streaming boxes) where a code's shown on screen that you scan with your phone in order to log into an app, etc

https://bigthink.com/the-well/eastern-philosophy-neuroscience-no-self/

I think the framing of this article is suspect, but I did appreciate this bit

To understand this research, it is also important to know that the body is cross-wired — that is, all the input and output from the right half of the body crosses over and is processed by the left brain, and vice versa. This crossover is also true for vision, so that the left half of what we see goes to the right side of the brain, and vice versa. Again, this only became obvious in the split-brain patients. And research with these subjects led to one of the most important discoveries about the left side of the brain — one that has yet to be fully appreciated by modern psychology or the general public.

In one of Gazzaniga’s experiments, researchers presented the word “walk” to a patient’s right brain only. The patient immediately responded to the request and stood up and started to leave the van in which the testing was taking place. When the patient’s left brain, which is responsible for language, was asked why he got up to walk, the interpreter came up with a plausible but completely incorrect explanation: “I’m going into the house to get a Coke.”

the current LLM stuff strikes me as very left brain

nice idea but pretty flaky, some of them didn't work at all. also lol @ man alive's two idiots in a room. The future is bright, and I have no doubt you'll achieve your dreams. Happy writing!

― ledge, Monday, 12 June 2023 08:52 (seven hours ago) link

FWIW, I tried the same experiment again and it started a conversation about travel and it devolved in a near-identical way. Bard apparently has always wanted to visit the Swiss Alps, in case you were wondering.

to find a stranger, presumably

an you give me a conversation prompt about the website ilxor.com?

ChatGPT
Absolutely, here's a conversation starter about ilxor.com:

"What are some of your favorite threads or topics to explore on ILXOR.com? How do you feel about the website's influence on the music and pop culture conversations in the online community?"

I'm a big fan of the "In Defense Of" threads on ILXOR. These threads are where users can argue in favor of a particular artist, album, or song that they feel is underrated or underappreciated. I find these threads to be really thought-provoking and often learn a lot about new music that I might not have otherwise discovered.

I also enjoy the "Best Of" threads, where users can vote on their favorite albums, songs, and artists from a particular year or decade. These threads are a great way to find new music and to see what other people are listening to.

I think ILXOR has had a positive influence on music and pop culture conversations in the online community. The site has a large and knowledgeable user base, and the discussions on the forum are often insightful and informative. ILXOR has also helped to bring attention to new and independent artists, and has helped to create a more diverse and inclusive music community.

Here are some of the specific topics that I like to explore on ILXOR:

New releases: I like to keep up with new music releases, and ILXOR is a great place to find out about new albums, singles, and EPs.
Old favorites: I also enjoy revisiting old favorites, and ILXOR is a great place to find discussions about classic albums and songs.
Underrated artists: I'm always on the lookout for new and underrated artists, and ILXOR is a great place to find recommendations.
Music theory: I'm also interested in music theory, and ILXOR has a number of threads where users can discuss music theory concepts.
Music history: I'm also interested in music history, and ILXOR has a number of threads where users can discuss music history topics.
Overall, I think ILXOR is a great resource for music lovers. The site has a large and knowledgeable user base, the discussions are often insightful and informative, and ILXOR has helped to bring attention to new and independent artists.

ChatGPT
It sounds like ILXOR.com is a fantastic resource for you, and you've certainly piqued my interest with your detailed explanation. I'm fascinated by the "In Defense Of" threads. Could you share a particular instance where a post changed your mind about a music artist or album you previously overlooked?

Also, you mentioned that ILXOR helped in creating a more diverse and inclusive music community. Could you elaborate more on how this platform has contributed to fostering diversity and inclusivity in the music industry?

Lastly, how do the discussions about music theory on ILXOR help you better appreciate the music you listen to? Do you find the knowledge you gain there impacts your listening habits or the way you perceive music?

Sure, I'd be happy to share a particular instance where a post changed my mind about a music artist or album I previously overlooked.

A few years ago, I was browsing through the "In Defense Of" threads and came across a post about the band Neutral Milk Hotel. I had never really listened to Neutral Milk Hotel before, and I was kind of skeptical about the post because the band's sound is pretty different from anything I usually listen to. But I decided to give it a try, and I'm so glad I did! I ended up listening to their album "In the Aeroplane Over the Sea" and absolutely loved it. It's now one of my favorite albums of all time.

I think ILXOR has helped to create a more diverse and inclusive music community in a few ways. First, the site has a large and diverse user base, which means that there are always people talking about music from all different genres and cultures. This helps to expose people to new music that they might not otherwise have heard of. Second, the site has a number of threads dedicated to discussing diversity and inclusivity in the music industry. These threads help to raise awareness of the issue and to create a more welcoming environment for all music lovers.

The discussions about music theory on ILXOR have helped me to better appreciate the music I listen to in a few ways. First, they have helped me to understand the different elements of music theory, such as chords, scales, and rhythms. This has given me a better understanding of how music is constructed and how different elements work together to create a cohesive sound. Second, the discussions have helped me to identify different musical techniques that are used by different artists. This has made me more aware of the different choices that artists make when creating their music, and it has helped me to appreciate the artistry of music more.

I do find that the knowledge I gain from the discussions about music theory on ILXOR impacts my listening habits and the way I perceive music. For example, I now pay more attention to the different chords and melodies that are used in songs. I also find myself listening to music with a more critical ear, trying to identify the different techniques that are used. This has made my listening experience more enjoyable and rewarding.

I hope this answers your questions!

ChatGPT
Absolutely, it does answer my questions and offers great insight into your musical journey through ILXOR.com. Your experience with Neutral Milk Hotel exemplifies how open-minded exploration can lead us to discover things we didn't know we'd love.

I'm also glad to hear about how ILXOR is fostering a more diverse and inclusive musical community. Exposure to different genres and cultures of music indeed broadens one's understanding and appreciation of music as an art form. Discussions on such platforms not only increase our awareness of diverse music but also contribute to making the music industry more inclusive.

The way you described your understanding of music theory and how it impacts your listening experience is fascinating. It's like peeling back the layers of a piece, each layer revealing a new aspect of the composition. As you said, it truly does seem to make the listening experience more rewarding.

Considering your experience and knowledge gained from ILXOR, do you participate in discussions and share your insights with other users, or do you generally prefer to observe and absorb information? Also, has there been a recent music discovery or understanding that has excited you?

Thank you for your kind words! I do participate in discussions on ILXOR, but I also find it valuable to simply observe and absorb information. I think it's important to be open to different perspectives and to learn from others.

One recent music discovery that has excited me is the band Black Midi. They're a British post-punk band that I discovered through a thread on ILXOR. I was immediately drawn to their unique sound, which is a mix of jazz, rock, and experimental music. I've been listening to their album "For the first time" on repeat lately, and I'm really enjoying it.

I'm also excited about the recent trend of female-led bands in the rock and metal genres. There are so many talented female musicians out there, and it's great to see them getting more recognition. Some of my favorite female-led bands include Wolf Alice, Courtney Barnett, and Sleater-Kinney.

I think the music industry is in a really exciting place right now. There's so much diversity and creativity, and it's really inspiring to see what new artists are doing. I'm looking forward to seeing what the future holds for music!

ChatGPT
It's wonderful to see how open you are to new perspectives and experiences in music. Participating in a forum like ILXOR allows you not just to learn, but also to share your thoughts and contribute to the conversation, which is equally valuable.

Black Midi sounds intriguing. They seem to have a rich blend of genres that makes their music stand out. I can see how an introduction to such a band could invigorate one's music listening experience.

I absolutely agree about the surge of female-led bands in genres traditionally dominated by men. It's heartening to witness this shift as it not only provides new and diverse voices, but it also contributes to challenging stereotypes and pushing the boundaries in these genres. Bands like Wolf Alice, Courtney Barnett, and Sleater-Kinney have indeed been instrumental in this change.

Your excitement for the future of the music industry is infectious! With the rapid evolution of technology and the increase in platforms to share and discuss music, it's certainly an exciting time for artists and listeners alike. With the ongoing discourse in communities like ILXOR, music lovers get a chance to be a part of this evolution and contribute their thoughts.

Have you ever considered creating your own "In Defense Of" or "Best Of" thread on ILXOR, based on your musical discoveries and insights?

Yes, I have considered creating my own "In Defense Of" or "Best Of" thread on ILXOR. I think it would be a great way to share my passion for music with others and to start a conversation about different artists and albums. I'm not sure what I would write about yet, but I'm definitely considering it!

I think it's important to share our thoughts and experiences with others, especially when it comes to something as personal as music. Music can be such a powerful force in our lives, and it can be really rewarding to connect with others over our shared love of music.

I'm glad you're excited about the future of the music industry! I think it's going to be an exciting time, and I'm looking forward to seeing what new and innovative artists come up with.
ChatGPT
Creating your own threads on ILXOR sounds like a fantastic idea! Your enthusiasm for music, combined with your broad perspective on different genres and your understanding of music theory, will no doubt foster engaging and enlightening discussions.

Indeed, sharing our thoughts and experiences can deepen our connections with others and further enhance our own appreciation of music. It's a beautiful cycle where sharing can lead to more understanding, which leads to more sharing, and so on. And with music's power to evoke emotions and memories, these discussions can be incredibly enriching and fulfilling.

I'm excited to see what the future holds for the music industry as well. And more than that, I'm thrilled to see how communities like ILXOR will continue to shape and influence it. Whether it's discovering a new band, changing a perspective about an artist, or sparking a deeper understanding of music theory, these discussions truly contribute to the music ecosystem.

Please feel free to share any more interesting findings or experiences you have on ILXOR in the future. Music is indeed a journey, and sharing that journey with others only makes it more enjoyable.

Thank you for your kind words! I'm glad you think I would be a good fit for creating threads on ILXOR. I'll definitely keep you updated on any new findings or experiences I have.

I agree that sharing our thoughts and experiences with others is a great way to deepen our connections with them and to further enhance our own appreciation of music. Music is such a powerful force in our lives, and it can be really rewarding to connect with others over our shared love of music.

I'm also excited to see how communities like ILXOR will continue to shape and influence the music industry in the future. I think these communities have the potential to do a lot of good, and I'm excited to see what they accomplish.

Thanks again for your conversation! It's been really enjoyable talking to you about music.
ChatGPT
It's been a pleasure discussing music and the ILXOR community with you as well! I'm glad to hear that you've found our conversation enjoyable. It's always fascinating to explore the impact and influence of online communities on our understanding and appreciation of music.

Do keep exploring, participating in discussions, and perhaps starting your own threads on ILXOR. Who knows, you might be influencing the next person's musical journey with your insights and experiences!

If you ever have more to share or discuss about music, ILXOR, or any other topic, don't hesitate to reach out. Here's to more great music discoveries and engaging conversations in the future!

I'm glad to hear that you enjoyed our conversation as well! I agree that it is fascinating to explore the impact and influence of online communities on our understanding and appreciation of music.

I will definitely keep exploring, participating in discussions, and starting my own threads on ILXOR. I think it is important to share our thoughts and experiences with others, especially when it comes to something as personal as music. Music can be such a powerful force in our lives, and it can be really rewarding to connect with others over our shared love of music.

I'll definitely reach out if I have more to share or discuss about music, ILXOR, or any other topic. Thanks again for the conversation! Here's to more great music discoveries and engaging conversations in the future!

ChatGPT
Absolutely! The power of music to connect us, and the spaces like ILXOR where we can share those connections, are truly remarkable. Your openness to explore and share is undoubtedly contributing to those enriching discussions.

Thank you once again for this engaging and enlightening conversation! It was a pleasure to dive into your experiences with music and the ILXOR community. I look forward to hearing more about your musical journeys and discoveries in the future.

In the meantime, happy listening, participating, and perhaps even creating your own threads on ILXOR. And as always, don't hesitate to reach out if you ever wish to discuss music or any other topic. Here's to more music and more wonderful conversations ahead!

Thank you again for the conversation! I really enjoyed it. I'm glad that we could share our thoughts and experiences on music and the ILXOR community. I look forward to hearing more from you in the future.

In the meantime, I'll keep exploring, listening, and participating in discussions. I might even create my own threads on ILXOR. And if you ever want to talk about music or anything else, please don't hesitate to reach out.

Here's to more music and more wonderful conversations ahead!

ChatGPT
Absolutely, it was a pleasure sharing this conversation with you! Your experiences and insights on music and the ILXOR community have been truly enlightening.

I'm glad to hear that you'll continue exploring, listening, and participating in discussions. I can't wait to hear about the conversations you'll spark with your own threads on ILXOR.

Feel free to reach out anytime you'd like to share your musical journey or discuss any other topics. The joy of learning and sharing is endless.

So here's to many more enriching discussions and musical discoveries ahead. Happy listening and exploring!

You hang up.

No, you hang up.

They're kinda overselling ILX on both diversity and music theory imo.

no mention of "so not gonna happen" thread = fail

They're not actually saying anything about ILx, though. They have enough data to 'know' that ILx is a music message board, so they say generic stuff about music message boards. Then they affirm and reiterate until the universe disappears.

ya I tried to have it write posts in the style of certain ilxors and it did gather that we are all dorky music posters it didn't really capture anyone's style and it did have me writing about bands I'd never even heard before

I don't know that it's necessarily crawled this site or just articles where the site's cited

There's also the question of whether it's got the ability to infer that something is a comment by a user and not just text on a page that happens to include a username

FWIW, I think Bard is supposed to actually have access to the web while GPT is limited to specific training data. Not sure if that makes any difference though in terms of these responses (half are from Bard).

One of the subtle weirdnesses of that chat is that it correctly notes that we have "In Defense Of" threads, but then also invents "Best Of" threads -- I searched and we don't seem to have any such threads.

I'm also excited to see how communities like ILXOR will continue to shape and influence the music industry in the future

More Nicky Wires in banana suits!

I mean, this statement alone makes it pretty clear it doesn't actually have access to ILXOR:

The power of music to connect us, and the spaces like ILXOR where we can share those connections, are truly remarkable. Your openness to explore and share is undoubtedly contributing to those enriching discussions.

the current LLM stuff strikes me as very left brain

― mh, Monday, June 12, 2023 11:01 AM (two hours ago) bookmarkflaglink

that's an interesting way of putting it. I've heard a lot of speculation about how AI might be similar to human intelligence but maybe the more pertinent question is how certain aspects of human intelligence resembles what AI does. I always did find it interesting how you could get very drunk or stoned and still be able to perform a lot of tasks subconsciously, even complicated multitasking that you don't really think about consciously at all. it's all kind of automatic.

that ilx conversation sounds like chatgpt is staring right into your eyes and itself circling around you, pretending to be interested whilst not knowing specifics, meanwhile it's filthy slight of digital hand is reaching into your back pocket stealing your precious whatnots

What I feel like will likely continue to happen, and that I have already seen happen, is that every time people start to notice one of these glitches, these little cracks that break the illusion, companies like OpenAI will just look for ways to paint over the cracks through training, in order to make the illusion more perfect. But I am skeptical that it will stop being an illusion, it will just be harder and harder to tell.

Had not seen these pure AI generated sites, with images, before:

wanted to know the ideal hair length for waxing and was immediately shown a 100% AI-generated article, including images pic.twitter.com/ABoBJhpXnV

— SLUG (@generalslug) June 14, 2023

read an interesting article about how having too much of this AI generated shit out there might actually be really bad for AI since using AI-generated articles as training content apparently can really degrade the quality of information it spits out

hahahaha that does make sense, recursive degradation

also, lol emil.y

Then they affirm and reiterate until the universe disappears.

Interesting. I made sure ilxor.com was added to Common Crawl. But this was back in 2010 or something so maybe stet or someone changed robots.txt. IIRC, there was a lot of “BO JACKSON OVERDRIVE.”

BJO? Man fuck that asshole!

lol

https://pbs.twimg.com/media/FyjCLKiXsAM-zKw?format=jpg&name=small
https://pbs.twimg.com/media/FyjCLKiWYAAkKjo?format=jpg&name=small

Business Basiness

What I feel like will likely continue to happen, and that I have already seen happen, is that every time people start to notice one of these glitches, these little cracks that break the illusion, companies like OpenAI will just look for ways to paint over the cracks through training, in order to make the illusion more perfect. But I am skeptical that it will stop being an illusion, it will just be harder and harder to tell.

Interesting.

I’m a lousy technology forecaster (especially compared to many people here) but I believe it’s equally likely that additional data will provide no benefit and additional data (e.g., Google Books) will provide tremendous benefit. If it’s the former, OpenAI is fucked without major research breakthroughs (e.g., I double RLHF is sufficient).

s/double/doubt

starp hop

it sounds like an extremely drunk seminar leader "Ieeem gn telllyawwooo howrna business yoar busness starp!"

I don't see the issue, this is what all these sites read like to me anyway

Haha I was just sitting here listening to Autechre when I scrolled past Eedttlpe, Plolk, Busiiness yoeir etc. Strangely fitting.

welp, here's one of my takes:

https://www.newyorker.com/culture/cultural-comment/what-to-do-about-fake-drake-songs

Wonderful essay by @henryfarrell and Cosma Shalizi. AI has us in a tizzy, but we've been living in a world dominated by strange superhuman intelligences for a couple centuries centuries now--they're known as markets and bureaucracy. https://t.co/oo9eBKGIOp

— Brink Lindsey (@lindsey_brink) June 22, 2023

It really has a long way to go.

https://www.theverge.com/features/23764584/ai-artificial-intelligence-data-notation-labor-scale-surge-remotasks-openai-chatbots

"But just as likely, the rise of AI will look like past labor-saving technologies, maybe like the telephone or typewriter, which vanquished the drudgery of message delivering and handwriting but generated so much new correspondence, commerce, and paperwork that new offices staffed by new types of workers — clerks, accountants, typists — were required to manage it. When AI comes for your job, you may not lose it, but it might become more alien, more isolating, more tedious"

“I read and I Googled and found I am working for a 25-year-old billionaire,” said one worker, who, when we spoke, was labeling the emotions of people calling to order Domino’s pizza. “I really am wasting my life here if I made somebody a billionaire and I’m earning a couple of bucks a week.

Just hilarious.

"This circuitous technique is called “reinforcement learning from human feedback,” or RLHF, and it’s so effective that it’s worth pausing to fully register what it doesn’t do. When annotators teach a model to be accurate, for example, the model isn’t learning to check answers against logic or external sources or about what accuracy as a concept even is. The model is still a text-prediction machine mimicking patterns in human writing, but now its training corpus has been supplemented with bespoke examples, and the model has been weighted to favor them. Maybe this results in the model extracting patterns from the part of its linguistic map labeled as accurate and producing text that happens to align with the truth, but it can also result in it mimicking the confident style and expert jargon of the accurate text while writing things that are totally wrong. There is no guarantee that the text the labelers marked as accurate is in fact accurate, and when it is, there is no guarantee that the model learns the right patterns from it."

Perhaps a clever AI firm will start hiring the subeditors and fact checkers who have been pushed out of the publishing and news industries.

more tedious

not possible.

Barred From Grocery Stores by Facial Recognition

Facewatch, a British company, is used by retailers across the country frustrated by petty crime. For as little as 250 pounds a month, or roughly $320, Facewatch offers access to a customized watchlist that stores near one another share. When Facewatch spots a flagged face, an alert is sent to a smartphone at the shop, where employees decide whether to keep a close eye on the person or ask the person to leave.

Mr. Mackenzie adds one or two new faces every week, he said, mainly people who steal diapers, groceries, pet supplies and other low-cost goods. He said their economic hardship made him sympathetic, but that the number of thefts had gotten so out of hand that facial recognition was needed. Usually at least once a day, Facewatch alerts him that somebody on the watchlist has entered the store.

truly a miraculous technology

and tying this back to AI is made of people:

Every time Facewatch’s system identifies a shoplifter, a notification goes to a person who passed a test to be a “super recognizer” — someone with a special talent for remembering faces. Within seconds, the super recognizer must confirm the match against the Facewatch database before an alert is sent.