6 ms·
What makes 5% of AI agents work in production?
- another_twist 1y agoSo I have read the MIT paper and the methodology as well as the conclusions are just something else. For example, the number comes from perceived successes and failures and not actual measurements. The customer conclusions are also - it doesnt improve or it doesnt remember. Literally buying into the hype of recursive self improvement and completely oblivious to the fact that API dont control model weights and such cant do much self improvement besides writing more CRUD layers. The other complaints are about integrations which are totally valid. But in industries which still run windows XYZ without any API platforms so thats not going away in those cases. Point being, if the paper itself is not very good discourse just a well marketed punditry, why should we discuss on the 5% number. It makes no sense.
- another_twist 1y agoIts weird that this makes the front page and Metas code world model never did.
- metadat 1y agoFirst I've heard of it: https://ai.meta.com/research/publications/cwm-an-open-weights-llm-for-research-on-code-generation-with-world-models/ https://ai.meta.com/research/publications/cwm-an-open-weight...
- CuriouslyC 1y agoHN front page dynamics are heavily driven by having readers of /new who are stans for your content.
- monero-xmr 1y agoA non-open ended path collapses into a decision tree. Very hard to think of customer support use-cases that do not collapse into decision trees. Most prompt engineering on the SaaS side results in very long prompts to re-invent decision trees and protect against edge cases. Ultimately the AI makes a “decision function call” which hits a decision tree. LLM is very poor replacement for a decision tree. I use LLM every day of my life to make myself highly productive. But I do not use LLM tools to replace my decision trees.
- LPisGood 1y agoIt just occurred to me that with those massive system files people use we’re basically reinventing expert systems of the past. Time is a flat circle, I suppose.
- LostMyLogin 1y agoAny chance you can ELI5 this to me?
- schrodinger 1y agoA decision tree is simply a model where you follow branches and make a decision at each point. Like... If we had tech support for a toaster, you might see: if toaster toasts the bread: if no: has turning it off and on again worked? if yes: great! you found a solution if no: hmm, try ... if yes: is the bread burnt after? if no: sounds like your toaster is fine! if yes: have you tried adjusting the darkness knob? if no: ship it in for repair if yes: try replacing the timer. does that help? if no: ship it in for repair if yes: yay you're toaster is fixed
- jongjong 1y agoIt's interesting because my management philosophy when delegating work has been to always start by telling people what my intent is, so that they don't get too caught up in a specific approach. Many problems require out-of-the-box thinking. This is really about providing context. Context engineering is basically a management skill. Without context, even the brightest people will not be able to fill in the gaps in your requirements. Context is not just nice-to-have, it's a necessity when dealing with both humans and machines. I suspect that people who are good engineering managers will also be good at 'vibe coding'.
- HardCodedBias 1y ago"I suspect that people who are good engineering managers will also be good at 'vibe coding'." I have observed that those who have both technical and management experience seem to be more adept (or perhaps willing?) to use LLMs in the daily life to good effect. Of course what really helps, like in all things, is conscientiousness and an obsession for working through problems (if people don't like obsession then tenacity and diligence).
- sbierwagen 1y ago>This Monday, I moderated a panel in San Francisco with engineers and ML leads from Uber, WisdomAI, EvenUp, and Datastrato. The event, Beyond the Prompt, drew 600+ registrants, mostly founders, engineers, and early AI product builders. >We weren’t there to rehash prompt engineering tips. >We talked about context engineering, inference stack design, and what it takes to scale agentic systems inside enterprise environments. If “prompting” is the tip of the iceberg, this panel dove into the cold, complex mass underneath: context selection, semantic layers, memory orchestration, governance, and multi-model routing. I bet those four people love that the moderator took a couple notes and then asked ChatGPT to write a blog post. As always, the number one tell of LLM output, besides the tone, is that by default it will never include links in the body of the post.
- stingraycharles 1y agoYeah, “here’s the reality check:”, “not because they’re flashy, but because they’re blah blah”. Why can’t anyone be bothered anymore to write actual content, especially when writing about AI, where your whole audience is probably already exposed to these patterns in content day in, day out? It comes off as so cheap.
- mccoyb 1y agoIt comes off as someone who lives their life according to quantity, not quality. The real insight: have some fucking pride in what you make, be it a blog post, or a piece of software.
- palmotea 1y ago> The real insight: have some fucking pride in what you make, be it a blog post, or a piece of software. The businessmen's job will be complete when they've totally eliminated all pride from work.
- philipallstar 1y agoThis same instinct is why a pencil costs almost nothing and is perfect, and isn't rubbish, really expensive, and created by someone who took pride in their work.
- AdieuToLogic 1y agoIt's funny that what the author identifies as "the reality check": Here’s the reality check: One panelist mentioned that 95% of AI agent deployments fail in production. Not because the models aren’t smart enough, but because the scaffolding around them, context engineering, security, memory design, isn’t there yet. Could be a reasonable definition of "understanding the problem to solve." In other words, everything identified as what "the scaffolding" needs is what qualified people provide when delivering solutions to problems people want solved.
- whatever1 1y agoThey fail because the “scaffolding” is building the complicated expert system that AI promised that one would not have to do. If I implement myself a strict parser and an output post-processor to guard against hallucinations, I have done 100% of the business related logic. I can skip the LLM in the middle altogether.
- nylonstrung 1y agoVery interesting that you've found ways to mitigate the hallucination issue. Are you able to share more about what worked for you with the post processor and parser?
- AdieuToLogic 1y ago> If I implement myself a strict parser and an output post-processor to guard against hallucinations, I have done 100% of the business related logic. I can skip the LLM in the middles altogether. Well said and I could not agree more.
- moduspol 1y ago> If I implement myself a strict parser and an output post-processor to guard against hallucinations, I have done 100% of the business related logic. I can skip the LLM in the middle altogether. You might even be able to put a UI on it that is a lot more effective than asking the user to type text into a box.
- danieltanfh95 1y ago
- ares623 1y agoAt some point, say 5 years from now, someone will revisit their AI-powered production workloads and ask the question "how can we optimize this by falling back to non-AI workload?". Where does that leave AI companies when the obvious choice is to do away with their services once their customers reach a threshold?
- EdwardDiego 1y ago"Huh, turns out we could replace it all with a 4 line Perl script doing linear regression."
- anonzzzies 1y agoA lot of what we encounter is; there is this 'chat' interface which is the 'wow factor': you type something in english and something (like text to sql) falls out, maybe 60-80% of what was needed. But then the frustration (for the user) starts: the finetuning of the result. After a few uses, they always ask for the 'old way' back to do that: just editing the query or give them knobs to turn to finetune the result. Where most want knobs which are, outside the most generic cases (pick a timespan for a datetime column), custom work. So AI is used for the first 10% of the work time (which gives you 60%+ of the solution) until the frustration lands: the last 40% or less are going to take 90% of your time. Still great as overall it will probably take far less time than before.
- hshdhdhehd 1y agoBase models are the seed, fine tuning is the genetically modified seed. Context is the fertiliser.
- handfuloflight 1y agoAgents are the oxen pulling the plow through the seasons... turning over ground, following furrows, adapting to terrain. RAG is the irrigation system. Prompts are the farmer's instructions. And the harvest? That depends on how well you understood what you were trying to grow.
- nylonstrung 1y agoLocusts are when LLMs inexplicably rewrite your existing code despite in-line and prompt instructions not to
- ath3nd 1y ago[dead]
- iagooar 1y agoWow, half of this article deeply resonates with what I am working on. Text-to-SQL is the funniest example. It seems to be the "hello world" of agentic use in enterprise environments. It looks so easy, so clear, so straight-forward. But just because the concept is easy to grasp (LLMs are great at generating markup or code, so let's have them translate natural language to SQL) doesn't mean it is easy to get right. I have spent the past 3 months building a solution that actually bridges the stochastic nature of AI agents and the need for deterministic queries. And boy oh boy is that rabbit hole deep.
- jamesblonde 1y agoText2SQL was 75% on bird-bench 6 months ago. Now it's 80%. Humans are still at 90+%. We're not quite there yet. I suspect text-to-sql needs a lot of intermediate state and composition of abstractions, which vanilla attention is not great at. https://bird-bench.github.io/ https://bird-bench.github.io/
- ares623 1y agoText to sql is solved by having good UX and a reasonable team that’s in touch with the customers needs. A user having to come up with novel queries all the time to warrant text 2 sql is a failure of product design.
- caust1c 1y agoThis is exactly it. AI is sniffing out the good datamodels from the bad. Easy to understand? AI can understand it too! Complex business mess with endless technical debt? Not too much. But this is precisely why we're seeing startups build insane things fast while well established companies are still questioning if it's even worth it or not.
- strange_quark 1y agoThis 1000x. I’ve sat through several vendor demos of BI tools that have a chatbot and seen my PM go all starry eyed that you can ask it “show me top x over the last week” and get a chart back. How an empty text box is easier to use than a UI with several filter drop-downs, I’ll never understand, and I suspect that the people impressed with this stuff don’t know either.
- hn_throwaway_99 1y ago> Here’s the reality check: One panelist mentioned that 95% of AI agent deployments fail in production. Not because the models aren’t smart enough, but because the scaffolding around them, context engineering, security, memory design, isn’t there yet. It's a big pet peeve of mine when an author states an opinion, with no evidence, as some kind of axiom. I think there is plenty of evidence that "the models aren't smart enough". Or to put it more accurately, it's an incredibly difficult problem to get a big productivity gain when an automated system is blatantly wrong ~1% of the time but when those wrong answers are inherently designed to look like right answers as much as possible.
- ath3nd 1y ago[dead]
- EdwardDiego 1y ago> One team suggested that instead of text-to-SQL, we should build semantic business logic layers, “show me Q4 revenue” should map to a verified calculation, not raw SQL generation. Okay, how would that work though? Verified by who and calculated by what? I need deets.
- tirumaraiselvan 1y agoA simple way is perhaps implement a text-to-metrics system where metrics could be defined as SQL functions.
- moomoo11 1y agopsychedelics
- dchftcs 1y agoOn one side, you have an agent calculating the revenue. On the other side, you have an SQL that calculates the revenue Compare the two. If the two disagree, get the AI to try again. If the AI is still wrong after 10 tries, just use the SQL output.
- mnky9800n 1y agoso you have an answer and then you throw compute at trying to produce the answer in a different way. What I hear is a billion dollar AI startup in the making!
- encyclopedism 1y agoWell put...lol :-)
- meheleventyone 1y agoThey're saying that someone should implement the CalculateQuarterRevenue(year, quarter) function somewhere in a manner that has been verified (e.g. run it against previous quarters to make sure it works correctly) then rather than using the LLM to generate SQL you use it to decide what domain function should be called. Which to me seems to mean that someone on the panel was gently taking the piss out of the idea. Since if you've done all the hardwork anyway presenting this in a deterministic way with a nice UX is straightforward bit of front end work.
- tirumaraiselvan 1y agoThis article is getting a lot of hate but honestly it does have good amount of useful content learned through practical experience, although at an abstract level. For example, this section: ``` The teams that succeed don’t just throw SQL schemas at the model. They build: Business glossaries and term mappings Query templates with constraints Validation layers that catch semantic errors before execution ``` Unfortunately, the mixing of fluffy tone and high level ideas is bound to be detested by hands on practitioners.
- intended 1y agoI just refuse to read long AI generated text. Sadly this feels exactly like that.
- codyb 1y agoI get really frustrated when I see it on PRs cause it's such a time sink, super obvious, and so fluffy. So you scaffold this up in 30 seconds but want me to read through it carefully? Cool, thanks.
- donnaoana 1y agoI am curious, how would you use AI then, if not to make one more productive? The text is not AI generated, I came up with the questions, moderated the discussion, wrote a draft that the speakers red, added comments and the AI polished it, the AI was a custom GPT that I trained on my previous text from that substack. I am curious what would you have done differently or if you would refuse to use AI at all? I wrote the article, so I am genuinly curious. I didn't know someone posted on Hacker News, I knew people like to be negative here because there is no accountablity, I want to learn from all this hate. I am personally happy with the outcome, I gover over 30 notes from people who are building that this was useful to them, and the speakers were happy. So I am curious what could have I done differently from your perspective or what should be my learning from all these people who take time from their day to hate on this piece of writing instead of deciding not to read and moving on.
- intended 1y agoHey it’s your call! As you said it’s your productivity. If you said it’s something you made for perusal and reading? Then it reads like AI. I’ve had to read tons of papers and articles, the most testing being conference submissions. I won’t read something with that structure unless I have to.
- thisisit 1y agoIt seems to me that people think AI is somehow magic. Recently I led a product demo. The conversation went something like this: End users (at my company) - Can your AI system look at numbers and find differences and generate a text description? Pre-sales - (trying to clarify) For our systems to generate text it will be better if you give it some live examples so that it understands what text to generate. End users - But there is supporting data (metadata) around the numbers. Can't your AI system just generate text? Pre-Sales - It can but you need to provide context and examples. Otherwise it is going to generic text like "there is x difference". End user - You mean I need to write comments manually first? That is too much work. Now these users have a call with another product - MS Copilot.
- alansaber 1y agoTBF synthetic data generation exists for this reason. I do understand why a lot of companies go with the "safe" choice (copilot) even though it's crap.
- alganet 1y ago> It seems to me that people think AI is somehow magic. That's because it is marketed as magic. It's marketed as magic so people will adopt the thing before knowing its shortcomings. https://pbfcomics.com/comics/the-masculator/ https://pbfcomics.com/comics/the-masculator/
- beezlebroxxxxxx 1y agoWell, you hear a lot about how AI will "empower" employees and generate new "insights" based off of data for analysts and execs. In reality, most executives aren't really interested in that. They'd like it for sure, but really what they want is automation. They want "efficiencies"; they want cost cutting. Anyone that's been involved in data science roles in corporate environments knows that "the data" is usually forced into an execs pre-existing understanding of a phenomenon. With AI, execs are really excited at "cutting out the middlemen" when the middlemen in the equation are very often their own paid employees. That's all fine and dandy in an abstract economic view, but it's sure something they won't say publicly (at least most won't). In terms of potential cost cutting, it probably is the most recent "new magic". You used to have to pay a consultant, now you can "ask AI".
- LogicFailsMe 1y ago95% of the talent is being paid top dollar to build ~5% of the applications?
- alansaber 1y agoAbsolutely, when we're talking about infrastructure versus model development (RL/fine tuning, let alone pre-training).
- janalsncm 1y ago> The panel’s consensus: conversation works when it removes a learning curve. Conversational UIs are controversial but I think there are a good number of websites where a better search could be more centric. Not generating text, but surfacing the most relevant text. I’m thinking of a lot of library documentation, government info websites, etc. Basically an improvement over deep hierarchical navigation, where their way of organizing info is a leaky abstraction. Maybe that will be one of the side effects of this AI boom. Who knows.
- marcosdumay 1y agoThose 5% that generate revenue on the MIT article do that because the only thing they are used for is creating marketing spam to send to people. And now we have an entire panel of bullshitters with an article-long theory about how to make LLMs program actually for real this time. (Oh, and it would be great if journalists actually cited their public sources, instead of pretending they link to the article but actually linking to their review of related content.)
- zoeey 1y agoI've always felt the real challenge isn't the LLM itself, but managing the context around it. Many people assume that writing a good prompt is enough, but the real work is turning something unpredictable into a tool you can actually rely on.