[HN Gopher] Good old-fashioned AI remains viable in spite of the...
       ___________________________________________________________________
        
       Good old-fashioned AI remains viable in spite of the rise of LLMs
        
       Author : webmaven
       Score  : 61 points
       Date   : 2023-12-02 16:29 UTC (6 hours ago)
        
 (HTM) web link (techcrunch.com)
 (TXT) w3m dump (techcrunch.com)
        
       | electroly wrote:
       | That is not what people typically mean by "GOFAI," which is a
       | term more commonly understood to refer to classical symbolic AI
       | as practiced by the MIT AI lab starting in the late 60s. From the
       | title I thought they were claiming that the OLD old fashioned AI
       | was still viable--they are not. They're still talking about
       | trained neural network models; they're distinguishing between
       | foundational models and single-task models.
        
         | randcraw wrote:
         | This. Symbolic AI (AKA GOFAI) died about 1985. But task-
         | oriented AI will never die, because the need for intelligent
         | agents with cutting-edge experise in niche skill areas costs
         | too much to embed all the available info on Earth into every
         | foundational AI model, much less, keep it up-to-date.
         | 
         | As the old saw says, the only perfect model of the universe is
         | the universe itself.
        
           | cjauvin wrote:
           | It didn't die per se I would say, it mainly got subsumed into
           | more general programming and problem solving patterns
           | (search, heuristics, A*, that sort of thing). What probably
           | "died" is the dream that those would be sufficient to achieve
           | general intelligence.
        
           | PartiallyTyped wrote:
           | "Symbolic" AI is very much used wherever you have theorem
           | provers e.g. z3, no?
        
             | tensor wrote:
             | Yes, and closely related are various forms of constraint
             | solvers, things that LLMS are still fantastically poor at.
        
               | PartiallyTyped wrote:
               | SMT Solvers are used extensively in certain parts of the
               | industry, especially when working in security-mindful
               | fields.
        
           | horsh1 wrote:
           | Nothing existing only before 1985, that is, before the
           | invention of the open source, was alive in the first place.
           | The code for the classic programs is nowhere to be found. And
           | if found, it is written in some interlisp dead language.
        
         | cscurmudgeon wrote:
         | > OLD old fashioned AI was still viable--they are not
         | 
         | False. Theorem provers exist and are widely used, sometimes
         | even in deep learning.
         | 
         | https://arxiv.org/pdf/2304.10558.pdf
        
         | tensor wrote:
         | Good old linear classifiers still solve many tasks close to
         | optimally. Probabilistic AI is not just neural networks.
        
       | minimaxir wrote:
       | As with most software development, modern AI work is all about
       | knowing your tools and when it's appropriate to use them. If you
       | have tabular data, even good LLMs will have trouble beating a
       | gradient-boost tree algorithm, but if you're working with
       | anything involving text data, you'll save yourself a lot of
       | hassle and likely get better results if you go directly to a
       | pretrained-LLM-generated text embeddings.
        
         | cuuupid wrote:
         | I've found the best LLMs still fall a bit short when it comes
         | to being truly best in class, SOTA in most useful tasks is
         | still dominated by smaller models with focused approaches and
         | innovative loss mechanisms.
        
         | tensor wrote:
         | Using fasttext as a first pass is probably the best idea for
         | production work.
        
           | minimaxir wrote:
           | fasttext/word2vec/training-your-own-word-embeddings-from-
           | scratch requires very optimistic assumptions about your data
           | that rarely hold well in practice, such as not having
           | idiosyncratic syntax from user input. I've learnt this the
           | hard way.
           | 
           | Transformers-based LLMs with a BPE tokenizer compensate for
           | this well.
        
         | miven wrote:
         | Well, LLMs don't consistently beat purpose-trained BERT-type
         | models just yet, for example in [1] the authors show RoBERTa
         | (fine-tuned for each individual task separately of course)
         | going basically toe-to-toe with GPT-4 on quite a few somewhat
         | conventional NLP tasks, while open-source LLaMA 2 models are
         | getting severely outclassed.
         | 
         | [1] https://arxiv.org/abs/2308.10092
        
           | minimaxir wrote:
           | You'll get far more than 80% of the way with just running a
           | pretrained text embeddings model.
           | 
           | There's always room for optimization (e.g. finetuning on your
           | own data) but it's an incredible baseline.
        
         | specproc wrote:
         | Yeah, embedding are where it's at for text. Most problems are
         | really about clustering or classification, when you like, get
         | down to it, man. Semantic similarity is one hell of a drug.
        
       | synthc wrote:
       | There are still tons of domains were GOFAI is very useful, like
       | planning and scheduling, and also in cases where you don't have
       | access to lots of training data or pretrained models.
        
       | bugglebeetle wrote:
       | If anything, LLMs just heighten their abilities. I can generate
       | synthetic datasets of size and quality for training BERT models
       | and the like that would've taken ages before. Similarly for
       | knowledge distillation from large models to small. It's never
       | been more exciting times for ML.
        
       | mejutoco wrote:
       | Also. Good oldfashioned statistics remains viable in spite of AI.
        
       | yeldarb wrote:
       | We've been seeing this in vision too. I'm really excited about
       | multimodal LLMs but it seems like they're going to be most useful
       | for solving new types of (qualitative) problems vs taking over
       | (quantitative) problems where small, fast CNNs have excelled.
       | 
       | Though they can also play well together[1] given _creating_
       | quantitative models is a qualitative problem.
       | 
       | [1] https://github.com/autodistill/autodistill
        
         | godelski wrote:
         | > We've been seeing this in vision too.
         | 
         | It's always amazing to me that __researchers__ still ask if
         | there are uses for GANs, thinking diffusion killed them. I have
         | a hard time taking anyone's claim that they are an expert in
         | synthetic image synthesis but it seems top companies hire them.
         | I see the same thing with ResNets and I just don't get it.
         | There's a tendency to not just railroad, but to actively build
         | it.
        
       | phillipcarter wrote:
       | One of the most common design patterns for using LLMs, Retrieval
       | Augmented Generation (RAG), is precisely a "good old fashioned
       | AI" problem to solve.
       | 
       | The quality of your application depends on the quality of your
       | data, how you organize it, how you understand results you get
       | from real-world usage, which model is used to compute semantic
       | embeddings, how you handle tricky retrieval problems (e.g., "show
       | me bananas" vs. "show me not bananas"), and so on.
       | 
       | All this means to me is that the possibilities around what can be
       | built is increasing, as are the needs for people who understand
       | both the "old fashioned" AI world and the new one that we're
       | stepping into.
        
       | gremgoth wrote:
       | Generative symbolic AI occurs all the time in data integration,
       | when you have to generate new identifiers (e.g. create new
       | targets for foreign keys that don't exist in any source),
       | recursively merge identified entities together along their
       | attributes, and so on. But now that this process is
       | mathematically well-understood it is no longer called "AI" but
       | rather "logic programming".
        
       | davesque wrote:
       | LLMs themselves are made up of "good old fashioned" components.
       | The public conversation about generative AI seems to assume that
       | it's doing something fundamentally different than what came
       | before. Of course, LLMs are still just deep NNs trained through
       | back propagation to "classify" strings of text into one of many
       | categories corresponding to each possible token. The only thing
       | they do differently is to compute auto-correlation coefficients
       | in each layer and zero some of them out. And of course they take
       | advantage of massive amounts of training data by self-
       | supervising.
       | 
       | LLMs are only a slight variation on the good old fashioned models
       | that already existed at the time they showed up.
        
         | cjauvin wrote:
         | What they bring in my mind that is deeply surprising is the
         | realization that from such a simple objective function
         | (predicting the next word) could emerge such a wide array of
         | general capabilities. Before LLMs, neural networks
         | accomplishing "narrow" tasks were understood to be powerful,
         | but not in such a general way.
        
           | rdedev wrote:
           | My understanding is that at some point the model needs to
           | gain an understanding of how the world works if it needs to
           | know the right set of words that come next to keep reducing
           | perplexity. But I don't think this can scale to AGI but I
           | have nothing to back it up
        
           | og_kalu wrote:
           | Predicting the next word as an objective is straightforward,
           | not simple. Think about what it would mean to predict
           | internet scale text accurately. To for instance, know the
           | results of scientific papers before they are uttered.
           | 
           | There's a difference between an objective and what you need
           | to do to fulfill it.
        
         | snarkconjecture wrote:
         | "Good old-fashioned AI" (GOFAI) is a term of art that refers to
         | symbolic AI, in contrast to deep NNs.
         | 
         | Maybe we need to rename it if deep learning is already "old
         | fashioned"!
         | 
         | (EDIT: to be clear, I have no idea what TechCrunch thinks GOFAI
         | means.)
        
           | davesque wrote:
           | Yeah, that's what I was trying to indicate with the
           | quotations. But thanks for adding the note.
        
           | DonHopkins wrote:
           | The AI Treadmill:
           | 
           | Looking backward: GOFAI (Good Old Fashioned AI) becomes BOFAI
           | (Bad Old Fashioned AI), then later becomes ROFAI (Retro Old
           | Fashioned AI) then eventually AOFAI (Antique Old Fashioned
           | AI)...
           | 
           | Looking forward: AI (Artificial Intelligence) progresses to
           | AGI (Artificial General Intelligence) progresses to AGIC
           | (Artificial General Intelligence Consciousness) progresses to
           | MAGIC (Miraculous Artificial General Intelligence
           | Consciousness)...
        
       | gdiamos wrote:
       | LLMs popularized zero-shot learning, or "prompt engineering"
       | which is drastically easier to use and more effective than
       | labeling data.
       | 
       | You can also retrofit "prompt engineering" onto good old fashion
       | ML like text classifiers. I wrote a library to do just that here:
       | https://github.com/lamini-ai/llm-classifier
       | 
       | IMO, it's a short matter of time before this takes over all of
       | what used to be called "deep learning".
       | 
       | Supervised learned, which required Herculean efforts to label big
       | datasets was important to prove that deep learning worked, but we
       | have been engineering it to be easier and more effective ever
       | since and there is no going back.
        
       | tedivm wrote:
       | One of my former coworkers at Rad AI always used to joke that
       | most problems people try to solve with AI could be solved better
       | with simple linear regression, and I really do think that input
       | was one of the reasons why that company is doing so well right
       | now. When you're trying to build a product you need to look at
       | the best tools to solve the problems you're working on. Sometimes
       | that truly is an LLM, and other times it isn't. Throwing a
       | technology into things just because it's the cool thing to do at
       | the moment is almost always a mistake though.
        
         | blowski wrote:
         | Following fashions can bring you a bit of short term cash, but
         | it won't build a long-term sustainable business.
        
         | chaostheory wrote:
         | That makes sense, but we often ignore and overlook the
         | marketing side for both attracting investors and customers. If
         | you need runway, it's "AI"
        
       | adastra22 wrote:
       | Article title is misleading. It is not about "Good old-fashioned
       | AI" (GOFAI) at all. That's a term of art referring to symbolic AI
       | like planning and explicit inference. All the stuff covered in
       | Norvig's classic "AI: A modern approach."
        
       ___________________________________________________________________
       (page generated 2023-12-02 23:02 UTC)