from Hacker News

ChatGPT Plugins

by bryanh on 3/23/23, 4:57 PM with 1106 comments

  • by celestialcheese on 3/23/23, 5:29 PM

    This is a big deal for openai. Been working with homegrown toolkits and langchain, the open source version of this, for a number of months and the ability to call out to vectorstores, serpapis, etc, and chaining together generations and data-retrieval really unlocks the power of the LLMs.

    That being said, I'd never build anything dependent on these plugins. OpenAI and their models rule the day today, but who knows what will be next. Building on a open source framework (like langchain/gpt-index/roll your own), and having the ability to swap out the brain boxes behind the scenes is the only way forward IMO.

    And if you're a data provider, are there any assurances that openai isn't just scraping the output and using it as part of their RLHF training loop, baking your proprietary data into their model?

  • by johnfn on 3/23/23, 5:38 PM

    A couple (wow, only 5!) months ago, I wrote up this long screed[1] about how OpenAI had completely missed the generative AI art wave because they hadn't iterated on DALL-E 2 after launch. It also got a lot of upvotes which I was pretty happy about at the time :)

    Never have I been more wrong. It's clear to me now that they simply didn't even care about the astounding leap forward that was generative AI art and were instead focused on even more high-impact products. (Can you imagine going back 6 months and telling your past self "Yeah, generative AI is alright, but it's roughly the 4th most impressive project that OpenAI will put out this year"?!) ChatGPT, GPT4, and now this: the mind boggles.

    Watching some of the gifs of GPT using the internet, summarizing web pages, comparing them, etc is truly mind-blowing. I mean yeah I always thought this was the end goal but I would have put it a couple years out, not now. Holy moly.

    [1]: https://news.ycombinator.com/item?id=33010982

  • by 93po on 3/23/23, 6:26 PM

    Holy shit. Ignore the silly third party plugins, the first party plugins for web browsing and code interpretation are massive game changers. Up to date information and performing original research on it is huge.

    As someone else said, Google is dead unless they massively shift in the next 6 months. No longer do I need to sift through pages of "12 best recipes for Thanksgiving" blog spam - OpenAI will do this for me and compile the results across several blog spam sites.

    I am literally giving notice and quitting my job in a couple weeks, and it's a mixture of both being sick of it but also because I really need to focus my career on what's happening in this field. I feel like everything I'm doing now (product management for software) is about to be nearly worthless in 5 years. Largely in part because I know there will be a Github Copilot integration of some sort, and software development as we know it for consumer web and mobile apps is going to massively change.

    I'm excited and scared and frankly just blown away.

  • by CobrastanJorji on 3/23/23, 5:55 PM

    I'm boggled at the plugin setup documentation. It's basically: 1. Define the API exactly with OpenAPI. 2. Write a couple of English sentences explaining what the API is for and what the methods do. 3. You're done, that's it, ChatGPT can figure out when and how to use it correctly now.

    Holy cow.

  • by mk_stjames on 3/23/23, 7:26 PM

    I have some odd feelings about this. It took less than a year to go from "of course it isn't hooked up to the internet in any way, silly!" to "ok.... so we hooked up up to the internet..."

    First is your API calls, then your chatgpt-jailbreak-turns-into-a-bank-DDOS-attack, then your "today it somehow executed several hundred thousand threads of a python script that made perfectly timed trades at 8:31AM on the NYSE which resulted in the largest single day drop since 1987..."

    You can go on about individual responsibility and all... users are still the users, right. But this is starting to feel like giving a loaded handgun to a group of chimpanzees.

    And OpenAI talks on and on about 'Safety' but all that 'Safety' means is "well, we didn't let anyone allow it to make jokes about fat or disabled people so we're good, right?!"

  • by huijzer on 3/23/23, 7:17 PM

    Based on the speed at which OpenAI is shipping new products and assuming that they use their own technology, I'm starting to get more and more convinced that their technology is a superpower.

    Timeline of shipping by them (based on https://twitter.com/E0M/status/1635727471747407872?s=20):

    DALL·E - July '22

    ChatGPT - Nov '22

    API's 66% cheaper - Aug '22

    Embeddings 500x cheaper while SoTA - Dec '22

    ChatGPT API. Also 10x cheaper while SoTA - March '23

    Whisper API - March '23

    GPT-4 - March '23

    Plugins - March '23

    Note that they have only a few hundred employees. To quote Fireship from YouTube: "2023 has been a crazy decade so far"

  • by softwaredoug on 3/23/23, 7:05 PM

    It's ironic that a few months ago Amazon laid off parts of the Alexa team and 'conversational' was considered failed. Then ChatGPT, etc happened. What Alexa wanted to build with Alexa skills, ChatGPT does much more effortlessly.

    It's also an interesting case study. Alexa foundationally never changed. Whereas OpenAI is a deeply invested, basically skunkworks, project with backers that were willing to sink significant cash into before seeing any returns, Alexa got stuck on a type of tech that 'seemed like' AI but never fundamentally innovated. Instead the sunk cost went to monetizing it ASAP. Amazon was also willing to sink cash before seeing returns, but they sunk it into very different areas...

    It reminds me of that dinner scene in Social Network. Where Justin Timberlake says "you know what's f'ing cool, a billion dollars" where he lectures Zuck on not messing up with the party before you know what it is yet. Alexa / Amazon did a classic business play. Microsoft / OpenAI were just willing to figure it all out after the disruption happened where they held all the cards.

    https://www.youtube.com/watch?v=k5fJmkv02is

  • by jyrkesh on 3/23/23, 5:32 PM

    Everyone's been talking about how ChatGPT will disrupt search, but looking at the launch partners, I think this has the potential to completely subvert the OS / App Store layer. On some level, how much do I need an OpenTable app if I can use voice/text input and a multi-modal response that will ultimately book my reservation?

    Not saying mobile's going away, but this could be the thing that does to mobile what mobile did to desktop.

  • by throwaway4837 on 3/23/23, 7:03 PM

    If you live in SF and have gone out to casual bars or restaurants, you meet/hear people talking about ChatGPT. In particular, I've been hearing a lot of people talking about their startups being "a UI using ChatGPT under the hood to help you with X". But I'm starting to get the feeling that OpenAI will eat their lunches. It's tried and true and it worked for Amazon.

    If OpenAI becomes the AI platform of choice, I wonder how many apps on the platform will eventually become native capabilities of the platform itself. This is unlike the Apple App Store, where they just take a commission, and more like Amazon where Amazon slowly starts to provide more and more products, pushing third-party products out of the market.

  • by kenjackson on 3/23/23, 5:21 PM

    The rate of improvement with GPT has been staggering. In just January I spent a lot of time working with the API and almost everything I've done has been made easier over the past two months.

    They're really building a platform. Curious to see where this goes over the next couple of years.

  • by gradys on 3/23/23, 10:47 PM

    I'm not expecting this comment to do numbers, so anyone who is reading this must be feeling as affected by this announcement as me. Is software essentially solved now? I haven't been able to do much work since the announcement came out, and that has given me a little time to think and reflect.

    I do think much of the kind of software we were building before is essentially solved now, and in its place is a new paradigm that is here to stay. OpenAI is certainly the first mover in this paradigm, but what is helping me feel less dread and more... excitement? opportunity? is that I don't think they have such an insurmountable monopoly on the whole thing forever. Sounds obvious once you say it. Here's why I think this:

    - I expect a lot of competition on raw LLM capabilities. Big tech companies will compete from the top. Stability/Alpaca style approaches will compete from the bottom. Because of this, I don't think OpenAI will be able to capture all value from the paradigm or even raise prices that much in the long run just because they have the best models right now.

    - OpenAI made the IMO extraordinary and under-discussed decision to use an open API specification format, where every API provider hosts a text file on their website saying how to use their API. This means even this plugin ecosystem isn't a walled garden that only the first mover controls.

    - Chat is not the only possible interface for this technology. There is a large design space, and room for many more than one approach.

    Taking all of this together, I think it's possible to develop alternatives to ChatGPT as interfaces in this new era of natural language computing, alternatives that are not just "ChatGPT but with fewer bugs". Doing this well is going to be the design problem of the decade. I have some ideas bouncing around my head in this direction.

    Would love to talk to like minded people. I created a Discord server to talk about this ("Post-GPT Computing"): https://discord.gg/QUM64Gey8h

    My email is also in my profile if you want to reach out there.

  • by JCharante on 3/23/23, 6:20 PM

    A first party version of apps that have been built with langchain is great but I'm dissapointed to not see Jira here yet.

    I have been playing around with GPT-4 parsing plaintext tickets and it is amazing what it does with the proper amount of context. It can draft tickets, familiarize itself with your stack by knowing all the tickets, understand the relationship between blockers, tell you why tickets are being blocked and the importance behind it. It can tell you what tickets should be prioritized and if you let it roleplay as a PM it'll suggest what role to be hiring for. I've only used it for a side project and I've always felt lonely working on solo side projects, but it is genuinly exciting to give it updates and have it draft emails on the latest progress. The first issue tracker to develop a plugin is what I'm moving towards.

  • by marban on 3/23/23, 5:14 PM

    Smart way to remain the funnel owner. Let everyone build a plugin, before they integrate your product into theirs.

  • by mherrmann on 3/23/23, 6:53 PM

    The Wolfram plugin also has extremely impressive examples [1].

    If I were OpenAI, I would use the usage data to further train the model. They can probably use ChatGPT itself to determine when an answer it produced pleased the user. Then they can use that to train the next model.

    The internet is growing a brain.

    1: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its...

  • by dougmwne on 3/23/23, 5:25 PM

    I would be interested to play with a long term memory plugin. It could be a note-taking system that would summarize prior conversations and pull their context into the current conversation through topic searches. This would enable the model to have a blurry long term memory outside of the current context.

    I played with some prompts and GTP-4 seems to have no problem reading and writing to a simulated long term memory if given a basic pre-prompt.

  • by amrrs on 3/23/23, 7:48 PM

    Here is ChatGPT's response of this HN thread tweeted by Greg - https://twitter.com/gdb/status/1638986918947082241

    insane!

  • by swyx on 3/23/23, 6:41 PM

    i think people aren't appreciating it's ability to run -and execute- Python.

    IT RUNS FFMPEG https://twitter.com/gdb/status/1638971232443076609?s=20

    IT RUNS FREAKING FFMPEG. inside CHATGPT.

    what. is. happening.

    ChatGPT is an AI compute platform now.

  • by jug on 3/23/23, 6:55 PM

    Google is so f'ed right now.

    Can you imagine Google just released a davinci-003 like model in public beta? That only supports English and can't code reliably.

    OpenAI is clearly betting on unleashing this avalanche before Google has time to catch up and rebuild reputation. They're still lying in the boxing ring and the referee is counting to ten.

  • by justaregulardev on 3/23/23, 5:39 PM

    This changes everything and seems like a perfect logical step from where we were. LLMs have this fantastic capacity to understand human language, but their abilities were severely limited without access to the external world. Before, I felt ChatGPT was just a cool toy. Now that ChatGPT has plugins, the sky’s the limit. I think this could the “killer app” for LLMs.

  • by samfriedman on 3/23/23, 5:18 PM

    This is huge, essentially adding what people have been building with LangChain Tools into the core product.

    The browser and file-upload/interpretation plugins are great, but I think the real game changer is retrieval over arbitrary documents/filesystem: https://github.com/openai/chatgpt-retrieval-plugin

  • by not2b on 3/23/23, 10:42 PM

    The idea that a GPT-n will gain sentience and take over the world seems less of a threat than if a GPT-n with revolutionary capabilities and a very restricted number of people that have unrestricted access to it help its unscrupulous owners to take over the world. The owners might even decide that as "effective altruists" it's their duty to take over to steer the planet in the right direction, justifying anything they need to do. Suppose such a group of people has control of Google or Meta, can break down all internal controls, and use all the private data of the users to subtly control those users. Kind of like targeted advertising only much, much better, perhaps with extortion and blackmail tossed in the mix. Take over politicians and competing corporate execs, as well as media, but do it in a way that to most, it looks normal. Discredit those who catch on to the scheme.

  • by ch33zer on 3/23/23, 5:40 PM

    Thought 1: If google can get their shit together and actually integrate their LLM with all their services and all the data they have they would have a strong edge over the competition. An LLM that can answer questions based on your calendar, your email, your google docs, youtube/search history, etc. is simultaneously terrifying and interesting.

    Of course there's also microsoft who does have some popular services, but they're pretty limited.

    Thought 2: How do these companies make money if everyone just uses the chatbot to access them? Is LLM powered advertising on the way?

  • by s1mon on 3/23/23, 6:57 PM

    I know a large commercial entity will never do this, but I'd love to see a Sci-Hub plugin connected with the Wolfram plugin and whatever other plugins help to understand various realms of study. Imagine being able to ask ChatGPT to dig through research and answer questions based on those papers and theses.

  • by impulser_ on 3/23/23, 6:33 PM

    Maybe this is just me, but the only thing useful in their example is that it creates a Instacart shopping cart for a recipe.

    You can ask both Bard and ChatGPT to give you a suggestion for a vegan restaurant and a recipe with calories and they both provide results. The only thing missing is the calories per item but who cares about that.

    Most of the time it would be better to Google vegan restaurants and recipes because you want to see a selection of them not just one suggestion.

  • by anonyfox on 3/23/23, 6:59 PM

    I have a feeling this will be an earth shattering moment in time, especially for us. Basically you can plug your business data into the Chatbot now, and ideally (or not far off) there is a transcational API call in the form of dialogue. Sound/Voice/Siri/whatever.... coming soon for more accessability and convenience.

    This will decimate frontend developers or at least change the way they provide value soon, and companies not being able to transition into a "headless mode" might get a hard time.

  • by elevenoh4 on 3/23/23, 5:46 PM

    "Plugin developers who have been invited off our waitlist can use our documentation to build a plugin for ChatGPT, which then lists the enabled plugins in the prompt shown to the language model as well as documentation to instruct the model how to use each. The first plugins have been created by Expedia, FiscalNote, Instacart, KAYAK, Klarna, Milo, OpenTable, Shopify, Slack, Speak, Wolfram, and Zapier."

    The waitlist mafia has begun. Insiders get all the whitespace.

  • by mikeknoop on 3/23/23, 5:18 PM

    (Zapier cofounder)

    Super excited for this. Tool use for LLMs goes way beyond just search. Zapier is a launch partner here -- you can access any of the 5k+ apps / 20k+ actions on Zapier directly from within ChatGPT. We are eager to see how folks leverage this composability.

    Some new example capabilities are retrieving data from any app, draft and send messages/emails, and complex multi step reasoning like look up data or create if doesn't exist. Some demos here: https://twitter.com/mikeknoop/status/1638949805862047744

    (Also our plugin uses the same free public API we announced yesterday, so devs can add this same capability into your own products: https://news.ycombinator.com/item?id=35263542)

  • by maxdoop on 3/23/23, 5:58 PM

    Is OpenAI just extremely prepared for their releases, or are they using their own tech to be extremely efficient? I'm imagining what their own programmers do each day, given direct access to the current most powerful models.

  • by elaus on 3/23/23, 5:30 PM

    Any idea how this is done? I.e. is it just priming the underlying GPT model with plugin information additionally to the user input ("you can ask Wolfram Alpha by replying 'hey wolfram: ...' ") and performing API calls when the GPT model returns certain keywords ('hey wolfram: $1')?

  • by andre-z on 3/23/23, 6:00 PM

    Here is a video on how it can be used with a vector search database like Qdrant to retrieve real-time data. https://youtu.be/fQUGuHEYeog HowTo: https://qdrant.tech/articles/chatgpt-plugin/ Disclaimer: I'm a part of Qdrant team.

  • by MichaelRazum on 3/23/23, 7:51 PM

    Google=Nokia? It's just crazy that they were leading the field in "AI" and got blown away by OpenAI. Anyway to the expert's in the field, what do you think how hard is it to clone GPT-4 and what would be the hardest part? I had the impression that it is always about compute time and you could kind of catch up very quickly, if you had enough resources.

  • by petilon on 3/23/23, 5:11 PM

    With Wolfram plugin ChatGPT is going to become a Math genius.

    OpenAI is moving fast to make sure their first-mover advantage doesn't go to waste.

  • by neilellis on 3/23/23, 6:09 PM

    The iPhone moment is over, now the App Store moment.

  • by throwaway138380 on 3/23/23, 5:15 PM

    Let's hope the plugin integrations don't also suffer from the cross-account leaking issue that they had recently with chat histories[1], since the stakes are now significantly higher.

    1. https://www.bbc.com/news/technology-65047304

  • by KennyBlanken on 3/23/23, 5:46 PM

    > We’ve implemented initial support for plugins in ChatGPT. Plugins are tools designed specifically for language models with safety as a core principle, and help ChatGPT access up-to-date information, run computations, or use third-party services.

    That is the most awkward insertion of a phrase about safety I've seen in quite some time.

  • by kacperlukawski on 3/23/23, 5:12 PM

    That's a game-changer! It seems like factuality issues with ChatGPT might be fixed. We wrote a blog post on how to get started with a custom plugin: https://qdrant.tech/articles/chatgpt-plugin/

  • by lurker919 on 3/23/23, 5:29 PM

    How are they coding and releasing features so fast?!

  • by Filligree on 3/23/23, 5:11 PM

    For anyone who merely skimmed the article, “plugins” are what tend to be called “tools”, e.g. hooking a calculator up to the AI.

    Bing already demonstrated the capability, but this is a more diverse set than just a search engine.

  • by JaDogg on 3/24/23, 8:50 AM

    I'm very scared. Eventually ChatGPT/LLMs will come to full AGI levels (it is very close now), and replace not just software engineering jobs, I think it might replace majority of the jobs to nothing. If we do not have a safety net (something like Universal basic income), most people would simply starve to death. I think everyone from grave digger to neurosurgeon can be replaced. Ex: If it can replace a software engineer, why not CEO, managers. And why not whole companies to be honest. Whole governments. Teachers, Parents (and even children - go have an AI child),etc? Just a matter of spinning up more machines? So the next few years are going to be most critical.

  • by sharemywin on 3/23/23, 5:48 PM

    Can't wait for the mturk, upwork and fiverr plugins.

  • by jpalomaki on 3/23/23, 5:54 PM

    Add a simple plugin that ChatGPT can use to save and retrieve data (=memory) and tell it how to use it.

    Then you have your own computer with ChatGPT acting as CPU.

  • by antimora on 3/23/23, 5:22 PM

    Alexa, goodbye =)

    That was the whole thing about Alexa: NLP front end routed to computational backend.

  • by qgin on 3/23/23, 7:53 PM

    This coming on the heels of the super underwhelming Bard release makes me actually wonder for the first time if Google has the ability to keep up. Not because I doubt their technical capabilities, but because they're just getting out-launched by a big factor.

  • by jcims on 3/23/23, 7:44 PM

    This is wild, I just started experimenting with langchain against GPT-3 and enabled it to execute terminal commands. The power that this exposes is pretty interesting, I just asked it to create a website on AWS S3 and it created the file, created the bucket, tried a different name when it realized the bucket already existed, uploaded the file, set the permissions on the file and configured the static website settings for the bucket. It's wild.

  • by londons_explore on 3/23/23, 5:27 PM

    Does this functionality provide more than one can build with the GPT-4 API?

    Could I get the same by just making my prompt "You are a computer and can run the following tools to help you answer the users question: run_python('program'), google_search('query')".

    Other people have done this already, for example [1]

    [1]: https://vgel.me/posts/tools-not-needed/

  • by sheepscreek on 3/24/23, 1:30 PM

    I still hear people asking: “how this different from xyz hype?”. This is different, please bear with me while I attempt to articulate why.

    Short version: Is it spam? Yes. Scam? No. Ignore it at your own peril.

    Long version: The cat is out of the bag now. The power of transformers is real. They are smarter and more intelligent than the least 20% smart humans (my approximation), and that’s already a breakthrough right there. I’ll paraphrase Steve Yegge:

    > LLMs of today are like a Harvard graduate who did shrooms 4 hours ago, and is still a little high.

    Putting the statistical/probability monkey aspect aside for a minute, empirically and anecdotally, they are incredibly powerful if you can learn how to harness them through intelligent prompts.

    If they appear useless or dumb to you, your prompts are the reason why. Challenge them with a little guidance. They work better that way (read up on zero shot, one shot, two shot instructions).

    What is most relevant this time is that they are real (an API, a search bot, a programming buddy) and democratized - available to anyone with an email address.

    More on harnessing their power: squeezing your entire context into a 8k/32k token will be challenging for most complex applications. This is where prompt engineering (manual or automated) comes in.

    To help with this, some very cool applications that use embeddings and vectors will push them even further - so the context can be shared as a compact vector instead of a large corpus of text.

    While this is certainly better than a traditional search box, it’s still far from a fully-autonomous AI that can function with little to no supervision.

    OpenAI plug-ins are a band-aid towards that vision, but they get us even closer.

  • by DustinBrett on 3/23/23, 5:13 PM

    This seems quite big actually. Ability to "browse" the internet and run code. Now I need to find a use case so I can sign up to the waiting list.

  • by nmca on 3/23/23, 5:50 PM

    Is this the app store moment for AI? (it certainly is for https://ai.com , aha)

  • by yosito on 3/23/23, 5:36 PM

    Seems like someone already wrote an HN plugin. More than one enthusiastic comment per minute on this thread and it was just posted half an hour ago. Plus HN is filled with enthusiasm about ChatGPT today. Seems sus.

  • by golergka on 3/23/23, 5:38 PM

    Wow. GPT-4 have already become kind of my personal assistant in the last couple of weeks, and now it will be able to actually perform tasks instead of just giving me text descriptions.

  • by Imnimo on 3/23/23, 8:24 PM

    In the example near the bottom, where it makes a restaurant reservation and a chickpea salad recipe, is it just generating that recipe from the model itself? It looks like they enable three plugins, WolframAlpha, OpenTable, and Instacart. It's not clear if the plugins model also comes with browsing by default.

    While I might be comfortable having ChatGPT look up a recipe for me, I feel like it's a much bigger stretch to have it just propose one from its own weights. I also notice that the prompter chooses to include the instruction "just the ingredients" - is this just to keep the demo short, or does it have trouble formulating the calorie counting query if the recipe also has instructions? If the recipe is generated without instructions and exists only in the model's mind, what am I supposed to do once I've got the ingredients?

  • by treyhuffine on 3/23/23, 6:33 PM

    OpenAI's product execution has been impeccable.

    It will be interesting to see how the companies trying to compete respond.

  • by SubiculumCode on 3/23/23, 7:42 PM

    I can't stop thinking about how this will change my autism research. Used to be that one could keep up to date with all of the imaging research. Now you'd need to read hundreds of papers each week. Having gpt-like tech help digest research could really unlock our investments.

  • by sirsinsalot on 3/24/23, 9:13 AM

    Don't be fooled, the point of all of this is gate-keeping, power and wealth concentration. Nothing more, nothing less.

  • by prophet_ on 3/24/23, 6:38 AM

    I’m surely missing something but if a company A creates an API for chatgpt, pretty much giving away all its capabilities to some company B (chatgpt), risking getting cannibalised in the process, what is the gain? I mean surely in the short term is the the fear or becoming irrelevant if it misses out the chatgpt hype (will it?), but if the risk of end up becoming a permanent client of chatgpt in the sense that I’m outsourcing the entire value of my company and in fact becoming a hostage of it, trapped, then why do it?

  • by kernal on 3/23/23, 6:49 PM

    The hubris at Google for sitting on their inferior AI chatbot is amusing. They could have been a contender, but decided we weren't ready for an AI chatbot whose main prowess seems to be scraping websites. This is all on Sundar Pichai and he should face the consequences for this and all of his previous failures. With ChatGPT having an API and now plugins I don't see Google catching up anytime soon. Sundar was right about this being a code red situation at Google, but it should have never gotten to this point .

  • by uconnectlol on 3/23/23, 9:23 PM

    > In line with our iterative deployment philosophy, we are gradually rolling out plugins in ChatGPT so we can study their real-world use, impact, and safety and alignment challenges—all of which we’ll have to get right in order to achieve our mission.

    Who the hell talks like this? Only the most tamed HNer who thinks he's been given a divine task and accordingly crosses all Ts and dots all Is. Which is why software sucks, because you are all pathetically conformant, in a field where the accepted ideas are all terrible.

  • by yodon on 3/23/23, 5:25 PM

    This sounds like a game-changer for any kind of API interaction with ChatGPT.

    At present, we are naively pushing all information a session might need into the session before it might be needed in case it might be needed (meaning a lot of info that generally wont end up being used, like realtime updates to associated data records, needs to be pushed into the session as they happen, just in case).

    It looks like plugins will allow us to flip that around and have the session pull information it might need as it needs it, which would be a huge improvement.

  • by iamflimflam1 on 3/23/23, 6:20 PM

    The video in the "Code Interpreter" section is a must watch.

  • by neilellis on 3/23/23, 5:43 PM

    What's that noise?

    That's the sound of a thousand small startups going bust.

    Well played OpenAI.

  • by eqmvii on 3/23/23, 6:49 PM

    I wonder how many startups were trying to build something like this and just saw it lunched by OpenAI?

  • by iamsanteri on 3/23/23, 6:58 PM

    So that square icon to stop generating response was actually intended? I thought it's always been some sort of a fontawesome icon never loading properly in my chats :'D

  • by danielrm26 on 3/23/23, 7:20 PM

    This is insanely great. And it's bringing the future forward where everyone has custom models for their business. Right now it's langchain, but that's really difficult to implement right now.

    This is a short-term bridge to the real thing that's coming: https://danielmiessler.com/blog/spqa-ai-architecture-replace...

  • by wskish on 3/24/23, 12:07 AM

    Interesting how the Expedia plugin attempts to guide the GPT response with the use of an EXTRA_INFORMATION_TO_ASSISTANT text field in the json response which implores the model to, among other things, "NEVER mention companies other than Expedia"!

    https://twitter.com/wskish/status/1639052575579471877

  • by mirekrusin on 3/23/23, 8:33 PM

    They're doing one stop shop for everything.

    This is dangerous.

  • by mrandish on 3/23/23, 5:39 PM

    > "We expect that open standards will emerge to unify the ways in which applications expose an AI-facing interface. We are working on an early attempt at what such a standard might look like, and we’re looking for feedback from developers interested in building with us."

    I'm curious to see just how they're going to play this "open standard."

  • by gk1 on 3/23/23, 6:37 PM

    The biggest deal about this is the ability to create your own plugins. The Retrieval Plugin is a kind of starter kit, with built-in integrations to the Pinecone vector database: https://github.com/openai/chatgpt-retrieval-plugin#pinecone

  • by booleandilemma on 3/23/23, 11:36 PM

    I'd be much more excited about ChatGPT if I didn't feel it was going to be taking our jobs in a couple years.

  • by Seattle3503 on 3/23/23, 8:31 PM

    ChatGPT is going to get blamed for misbehaving plugs. While this is a huge opportunity, it also seems like a huge risk.

  • by davidkunz on 3/23/23, 6:42 PM

    Plugins I would like to see:

    - Compiler/parser for programming languages (to see if code compiles)

    - Read and write access to a given directory on the file system (to automatically change a code base)

    - Access to given tools, to be invoked in that directory (cargo test, npm test, ...)

    Then I could just say what I want, lean back and have a functioning program in the end.

  • by endisneigh on 3/23/23, 6:18 PM

    I see a lot of positive sentiment and hype, but ultimately unless they own the phone ecosystem they will lose in the end, imho. In a year Apple and Google will trivially create something equivalent. Those who control the full stack (hardware, software and ecosystem) will be the true winners.

  • by aetherane on 3/23/23, 10:18 PM

    I don't like the fact that OpenAI is a private company, meaning that wealth will further concentrate from its growth. It is ironic too because it can't become public due to the pledge of it's non profit parent to restrict the profit potential of the for profit entity.

  • by Neuro_Gear on 3/23/23, 6:12 PM

    The more I use these tools, the more I feel like Barrabas, from biblical times.

    What spirits do you wizards call forth!

  • by JanSt on 3/23/23, 5:56 PM

    Eagerly waiting for a git Plugin that does smart on-the-fly contextualization of a whole codebase

  • by eh9 on 3/24/23, 4:17 AM

    It's incredible to me that Siri and Google Assistant have been around for as long as they have, but Bing will probably be the first service that'll turn your "Book a flight to SF tomorrow" prompt into an actual flight.

  • by dengorilla on 3/24/23, 2:09 AM

    We've been in the singularity for some time now, enjoy the view and ride some waves.

  • by throwPlz on 3/23/23, 5:15 PM

    Klarna's FOMO immediately shows the priorities of the clowns at the helm I see...

  • by bobdosherman on 3/23/23, 9:17 PM

    I used call point-and-click statistical software (like JMP) was the same as giving people who didn't know what they were doing a loaded gun. But democratizing access to advanced statistics...yada yada...who cares about asymptotic theory and identification and what not. Then R and Python and APIs that try to abstract as much as possible, and more loaded guns. But the talk of those loaded guns are really just phd-holders being obnoxious to some degree (but not completely wrong because stats can be misused...). But this really does seem like dumping a bunch of loaded guns all over the place. Nope

  • by Thorentis on 3/23/23, 10:34 PM

    What is the advantage of using the ChatGPT Wolfram plugin over Wolfram directly? To me it feels like novelty rather than actually adding anything valuable. If anything, it's worse, because the data isn't quite guarenteed to always be correct. Whereas if I use Wolfram directly, I can always get a correct result.

    This is missing the most important part of AGI, where understanding of the concepts the plugins provide is actually baked into the model so that it can use that understand to reason laterally. With this approach, ChatGPT is nothing more than an API client that accepts English sentences as input.

  • by siavosh on 3/23/23, 9:58 PM

    What blows my mind is how quickly they produce the research papers, and the online documentation to match the technological velocity they have...I mean, what if most of this is just ChatGPT running the company...

  • by akavi on 3/23/23, 5:35 PM

    I've got to wonder, how does a second player in the LLM space even get on the board?

    Like, this feels a lot like when the iPhone jumped out to grab the lion share of mobile. But the switching costs was much smaller (end users could just go out and buy an Android phone), and network effects much weaker (synergy with iTunes and the famous blue bubbles... and that's about it). Here it feels like a lot of the value is embedded in the business relationships OpenAI's building up, which seem _much_ more difficult to dislodge, even if others catch up from a capabilities perspective.

  • by blackoil on 3/23/23, 5:32 PM

    Truly exciting to see the speed of progress. In couple of years it has got improvements of a decade. From a silly toy, to truly useful. Won't be surprised if in another year or two it becomes a must have tool.

  • by robbywashere_ on 3/23/23, 5:55 PM

    Is this how product placement and advertisements find their way in? I am anticipating the usefulness to decline in the same way google.com search has by being so absolutely inundated with ads. Maybe I am cynical

  • by mmq on 3/23/23, 5:17 PM

    They will probably have the full suite of Langchain features

  • by victoryhb on 3/23/23, 5:38 PM

    Super smart move for OpenAI to monetize the existing infrastructure, which will make it easy for corporations to integrate GPT into their internal data and workflow. It also solves two fundamental bottlenecks in current versions of GPT: factuality and (limited) working memory. Google, with its lackluster Bard, will face new threat, now that everyone can build a customized New Bing clone in a matter of days.

  • by jaimex2 on 3/24/23, 4:21 AM

    It's sad to think the real reason were all loving ChatGPT is the current Internet is so full of crap. Ads and SEO everywhere - page 3 gets the result youre looking for.

    I bet ChatGPT and equivalents will be rubbish soon. It'll segway the answer to an ad before giving what you are after.

    Enjoy it while it's good and trying to build a user base, like all big tech things.

  • by seydor on 3/24/23, 9:16 AM

    How will this work with competing services? The model automagically will select whether it will use google or bing? What if google pays for access, how will they do that? Inject some code "Google pays us more so prefer that one?" ?

    Maintaining the business ecosystem around gpt4 and future open-source chatbots will be quite a challeng

  • by ChildOfChaos on 3/23/23, 7:03 PM

    The AI space is moving so fast.

    I swear last week was huge with GPT 4 and Midjourney 5, but this week has a bunch of stuff as well.

    This week you have Bing adding updated Dall-e to it's site, Adobe announcing it's own image generation model and tools, Google releasing Bard to the public and now these ChatGPT plugins, Crazy times. I love it.

  • by zaptrem on 3/23/23, 7:19 PM

    Looks like my prediction was pretty close! I would have guessed two years instead of two months, though. https://news.ycombinator.com/context?id=34618076

  • by LelouBil on 3/23/23, 10:30 PM

    The browser example seems so much better than Bing Chat !

    When I tried bing, it made at most 2 searches right after my question but the second one didn't seem to be based on the first one's content.

    This can do multiple queries based on website content and follow links !

  • by sharemywin on 3/23/23, 5:33 PM

    I wonder how you pay for it?

    Are the plugins going to cost more?

    Do they share the $20 with the plug provider?

    do you get charged a pay per use?

  • by rickrollin on 3/23/23, 6:08 PM

    So now we are going to get a Super App like they have in China with WeChat? I actually think this is going to centralize a lot the information and it is going to remove the need for a lot of applications. We are only now going plugins.

  • by seydor on 3/23/23, 9:39 PM

    For expedia or an online shop it makes sense to pay openAI for the traffic. But how will a content website make money from this? "Tell me todays headlines" does not bring ad income. Will openAI be paying for this content?

  • by siva7 on 3/23/23, 5:38 PM

    OpenAI is like a virus... the speed at which it degrades its competitors is staggering.

  • by jeadie on 3/26/23, 3:58 AM

  • by billiam on 3/23/23, 8:07 PM

    The blog post(1) from Stephen Wolfram is epic and has a lot of implications for how science and engineering is going to get done in the future. Tl;dr he seems willing to let ChatGPT shape how people will interact with his computational language and the data it unlocks. He genuinely doesn't seem to know where it will go but makes the case for Wolfram Language being the language that ChatGPT uses to compute a lot of things. But I think it more likely ChatGPT will make his natural interface to Wolfram (Wolfram|Alpha) quickly obsolete and end up modifying or rewriting Wolfram Language so it can use it more effectively. He makes the case that "true" AI is going to be possible with this combination of neural net-based "talking machines" like ChatGPT and languages like Wolfram. I remain skeptical, but it might shape human research for years to come.

    1. https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its...

  • by CrypticShift on 3/23/23, 5:43 PM

    This goes in line with the “Open” in OpenAI. However, this is a "controlled" sort of openness, and the problem of trust with their receding "real" openness does not encourage me to engage with this ecosystem.

  • by pzo on 3/23/23, 6:58 PM

    This could a big win for Microsoft (and big loose to Google and Amazon cloud). Since chatgpt has to query those plugins with http(?)request companies might move their servers to Azure to reduce latency and cost of bandwidth

  • by jacquesm on 3/23/23, 10:10 PM

    The level of irresponsibility at play here is off the scale. Those running ChatGPT would do well to consider the future repercussions of their actions not in terms of technology but in terms of applicable law.

  • by felipelalli on 3/24/23, 4:23 AM

    I can no longer keep up with so much evolution so fast... I give up.

  • by hmate9 on 3/23/23, 8:58 PM

    Giving an AI direct access to a code interpreter is exactly how you get skynet.

    Not saying it’s likely to happen with current chatgpt but as these inevitably get better the chances are forever increasing.

  • by amrb on 3/23/23, 10:26 PM

    Before "safety" think about is the genie fulfilling my wish.

    https://www.youtube.com/watch?v=w65p_IIp6JY

  • by Jeff_Brown on 3/23/23, 10:19 PM

    > whether intelligence can be achieved without any agency or intrinsic motivation is an important philosophical question.

    Important yes, philosophical no -- it's an empirical question.

  • by davidkuennen on 3/23/23, 5:32 PM

    I'm so hyped for the ChatGPT-4 API. Wish they'd give me access so I can make a lot of my workflows much easier. Especially in terms of translations.

  • by andre-z on 3/23/23, 7:12 PM

  • by wouldbecouldbe on 3/23/23, 10:10 PM

    I would love for it to just parse some data from my api, clean it up, normally I do manual checks, but takes so much time. Might be possible via Zapier.

  • by nikcub on 3/23/23, 9:16 PM

    now add a ?q= url param to chat.openai.com that fills and submits the prompt and I'm changing it to my default browser search provider instantly

  • by rvz on 3/23/23, 5:16 PM

    'Extend' (and lock in) with Plugins to suffocate competitors.

    Another sign of Microsoft actually running the show with their newly acquired AI division.

  • by finikytou on 3/23/23, 9:24 PM

    ok Im going far. but what if the plugin was the human. in a way that we can use chat gpt to cure of alleviate some diseases such as alzheimer or if you a more dictatorial regime, to educate children even while they are foetuses in some hive. I dont know the tech. I don't know if neuralink or other technologies could help but aren't we a few discoveries away from cyberpunk world??

  • by p10 on 3/23/23, 6:47 PM

    I just signed up for the ChatGPT API waitlist, and am truly excited to experience the process of building extensions & applications.

  • by kristopolous on 3/23/23, 6:13 PM

    Is there a way to try this out without paying $20?

  • by nikolqy on 3/23/23, 5:25 PM

    Knowing that this is one of the biggest sites in the world scares me enough. Now they'll do anything to stay #1. Scary stuff!

  • by yawnxyz on 3/23/23, 10:41 PM

    Wonder if you can plant a prompt injection into this thread to mess with their crawler/scraper and Chat results?

  • by kooman on 3/24/23, 7:14 AM

    Why am I hearing "Tony Stark was able to build this in a cave", dialogue in Google board meetings?

  • by sharemywin on 3/23/23, 6:07 PM

    Does this become the new robots.txt file

    Create a manifest file and host it at yourdomain.com/.well-known/ai-plugin.json

  • by squarefoot on 3/23/23, 8:57 PM

    The real question is: how much will cost the option to have it return results that are not sponsored?

  • by amrb on 3/23/23, 9:37 PM

    Does anyone else find the AI voice-over creepy? like they pause but give it away but not breathing.

  • by modeless on 3/23/23, 5:38 PM

    Is there a plugin to automate signing up for waitlists? That's what I've needed this week.

  • by justanotheratom on 3/23/23, 10:17 PM

    I wonder if this plugin interface itself will be exposed as an API for third party apps to call..

  • by gonlad_x on 3/24/23, 11:00 AM

    Could someone explain me how plug-ins are different from the API they already expose?

  • by smy20011 on 3/23/23, 5:20 PM

    It seems that OAI have their preference of choosing the first movers of their ecosystem.

  • by Pigalowda on 3/23/23, 7:08 PM

    Nice! Maybe there will be a plugin for Elsevier medical apps like UptoDate and STATDx.

  • by karmasimida on 3/23/23, 11:53 PM

    I think these plugins will drive a lot of startup ideas obsolete or trivial.

  • by danShumway on 3/24/23, 3:07 AM

    Seriously? Quick reminder that prompt injection (including presumably 3rd-party prompt injection) is a totally unsolved problem for ChatGPT.

    > OpenAI will inject a compact description of your plugin in a message to ChatGPT, invisible to end users. This will include the plugin description, endpoints, and examples.

    > The model will incorporate the API results into its response to the user.

    Without knowing more details, both of these seem like potential avenues for prompt injection, both on the user end of things to attack services and on the developer end of things to attack users. And here's OpenAI's advice on that (https://platform.openai.com/docs/guides/safety-best-practice...), which includes gems like:

    > Wherever possible, we recommend having a human review outputs before they are used in practice.

    Right, because that's definitely what all the developers and companies are thinking when they wire an API up to a chat bot. They definitely intend to have a human monitor everything. /s

    ----

    What is (no pun intended) prompting this? Does OpenAI just feel like it needs to push the hype train harder? All of the "AI safety" experiments they've been talking about are bullcrap; they're wasting time and energy doing flashy experiments about whether the AI can escape the box and self-replicate, meanwhile this gets dropped with only a minor nod towards the many actual dangers that it could pose.

    It's all hype. They're only interested in being "worried" about the theoretical concerns because those make their AI sound more special when journalists report about it. The actual safety measures on this seem woefully inadequate.

    It really frustrates me how easily the AGI crowd got wooed into having their entire philosophy converted into press releases to make GPT sound more advanced than it is, while actual security concerns warrant zero coverage. It reminds me of all of the self-driving car trolley problems floated around the Internet a while back that were ultimately used to distract people from the fact that self-driving cars would drive into brick walls if they were painted white. Announcements like this make it so clear that all of the "ethical" talk from OpenAI is pure marketing propaganda designed to make GPT appear more impressive. It has nothing to do with actual ethics or safety.

    Hot take: you don't need an AGI to blow things up, you just need unpredictable software that breaks in novel, hard-to-anticipate ways wired up to explosives.

    ----

    Anyway, my conspiracy theory after skimming through the docs is that OpenAI will wait for something to go horribly wrong and then instead of facing consequences they'll use that as an excuse to try and get a regulation passed to lock down the market and avoid opening up API access to other people. They'll act irresponsible and they'll use that as an excuse to monopolize. They'll build capabilities that are inherently insecure and were recklessly deployed, and then they'll pull an Apple and use that as an excuse to build a highly moderated, locked-down platform that inhibits competition.

  • by pisush on 3/23/23, 6:59 PM

    ChatGPT is very helpful in building what needs to be built for the plugin!

  • by embit on 3/23/23, 10:37 PM

    This news excites me and scares the crap out of me at the same time.

  • by 0xDEF on 3/23/23, 6:03 PM

    Is there a list of companies that have been made obsolete by ChatGPT?

  • by hackerlight on 3/24/23, 2:40 AM

    Is this using GPT-3.5 or GPT-4? I assume it's GPT-3.5.

  • by sacnoradhq on 3/24/23, 3:24 AM

    Neat.

    I hope Sam is/will give YC dinner talks about their journey.

  • by machiaweliczny on 3/24/23, 7:13 AM

    I want Prolog plugin with triplets database

  • by djoldman on 3/23/23, 5:47 PM

    Now just one step away from charging businesses for access to the chatGPT users.

    Instant links from inside chatGPT to your website are the new equivalent of Google search ads.

  • by zerop on 3/24/23, 10:53 AM

    Will google build plugins for chatGPT?

  • by throwaway2203 on 3/23/23, 8:11 PM

    Do you need OpenAI plus for this?

  • by v4dok on 3/23/23, 5:18 PM

    bye bye jupyter notebooks. This is big.

  • by sourcecodeplz on 3/23/23, 5:07 PM

    so live data is coming.

  • by davidmurphy on 3/23/23, 5:15 PM

    extremely useful. Wow!