[momojo]

This reminds me of Antirez's "Don't fall into the anti-AI hype" [0]

In a sentence: These foundation models are really good at optimizing these extremely high level, extremely well defined problem spaces (ie multiply matrices faster). In Antirez's case, it's "make Redis faster".

There have been two reactions: "Oh it would never work for me" and "I have seen months of my life accomplished in an hour", and I think they're both right. I think we should be excited for Antirez, (who has since been popping off [1]), and I think the rest of us should rest easy knowing that LLM's can't (and maybe were never meant to) tackle the tacit-knowledge-filled, human-system-centric, ambiguously-defined-problem-space jobs most mortals work.

[0] https://antirez.com/news/158 [1] https://antirez.com/news/164

[poisonfountain]

>I think the rest of us should rest easy knowing that LLM's can't (and maybe were never meant to) tackle the tacit-knowledge-filled, human-system-centric, ambiguously-defined-problem-space jobs most mortals work

I don't believe that anymore, to be honest. Models are starting to get good at ambiguity. Claude Code now asks me when something is ambiguous. Soon, all meetings will be recorded, transcribed and stored in a well-indexed place for the agents to search when faced with ambiguity (free startup idea here!). If they can ask you now, they'll be able to search for the answers themselves once that's possible. In fact, they already do it now if you have a well-documented Notion/Confluence, it's just that nobody has.

It's probably harder to RL for "identify ambiguity" than RL'ing for performance algorithms, sure, but it's not impossible and it's in the works. It's just a matter of time now.

[momojo]

> Models are starting to get good at ambiguity

That's fair, and something I've observed too. I wish I had written "the rest of us shouldn't freak out and quit software today".

But here's another data point: At the biotech I work for, writing good code has never been the bottleneck. I actually told my boss that a paid Claude vs free subscription wouldn't be that much value because even if it took every piece of code or algorithm we've ever written and 10x-ed the hell out of them, we'd still be bottlenecked by the biology and physics which dictates that we wait 24 days for our histology assay pipeline.

I have a hunch most fields outside of software are this way. And I'm personally not planning to quit anytime soon.

[poisonfountain]

Ok, but you job is clearly not a good sample for a "job most mortals work on".

[DoctorOetker]

Everybody likes to think their job or specialization is the bottleneck.

I think the domain knowledge and scientific knowledge is more often the bottleneck than people like to admit.

Sure it doesn't take a lot of knowledge to repeat or automate what the previous generations did, but such decisions were very far from the optimum considering the actual scientific frontier.

[momojo]

1. I admit my job is not a good sample of 'job most mortals work' 2. I still think 'most mortals' software devs aren't Antirez, and work in industries or fields that still require a human to bridge the ambiguity gap.

[TranquilMarmot]

> Soon, all meetings will be recorded, transcribed and stored in a well-indexed place for the agents to search when faced with ambiguity (free startup idea here!)

We were doing that over at Vowel a few years back, unfortunately it didn't pan out because you're competing directly against Zoom, Google Meet, Microsoft Teams, etc. They are all (slowly) catching up to where we were as a scrappy startup 4 years ago.

It was truly game-changing to have all of your meetings in an easily searchable database. Even as a human.

[dzhiurgis]

Why record when it can build in realtime as meeting is going on.

Slack is kinda there with Salesforce - can do a lot already on Agentforce and in Slackbot, but two aren't integrated just yet and Slackbot doesn't support group chats/channels. One interesting aspect in this will be - who has superiority boss, client, analyst or developer?

[risyachka]

In coding the ambiguity is very, very limited and constrained compared to any non dev job that involves any decision making

[exfalso]

That's.. not even close to being the case. It's literally a series of ambiguous questions and strategic decisions.

Non-ambiguous is like a first semester algorithms class in university.

[antonvs]

There seems to be a category of "coder" which fits the other commenter's description, where someone else makes all the significant decisions and they just write the code. Not coincidentally, that category seems most at risk from AI, because they're basically like a human version of a coding agent.

[mrdomino-]

Tacit knowledge is definitionally not recorded in any of these systems. This proposes to solve the problem of tacit knowledge by getting rid of it. It is not clear to me if that solution is either possible or desirable.

[nl]

The labs are spending hundreds of millions of dollars hiring people doing many fairly random (but economically valuable) tasks to collect this tacit knowledge for RL.

It works really well.

[mrdomino-]

It ceases to become tacit as soon as it is collected.

Maybe this rephrase will help: the proposed solution is to render all knowledge explicit.

[nl]

> It ceases to become tacit as soon as it is collected.

I'm not sure.

It it is collected via preferences then it isn't necessarily something that can be communicated (except in the LLM's latent space).

That still feels tacit to me.

To simplify that argument, the relationship between King and Queen in the Word2Vec latent space can be easily explicitly labelled.

But the relationship between Napoleon and Tsar Alexander I also exists and encodes much of the tacit knowledge about their relationship but isn't as easily labelled (eg, Google AI Mode says "Napoleon I and Tsar Alexander I had a volatile "bromance" that shifted from mutual admiration to deep animosity, acting as a defining conflict of the Napoleonic Wars".)

Word2Vec is a very simple model. In a more complex LLM that deeper knowledge can be queried by asking questions but you can never capture it all. Isn't that what "tacit knowledge" is?

[mrdomino-]

It's a good question, yeah, and a lot of these boundaries get fuzzy when they're looked at closely enough.

It's certainly the case that LLMs already are able to represent and make use of some kinds of apparently still tacit knowledge, and that the scope of that is apparently expanding. I don't question that. I question two things: whether it is always desirable for that scope to expand, and whether it is possible for that scope to ever fully cover what it seeks to cover.

[Yokohiii]

So self chosen total surveillance and transparency so your fav LLM can be better?

[DennisP]

Could always use a local LLM for stuff like that. One of my relatives works for one of the big audit firms and that's what they do.

[Yokohiii]

Sure. Still from what he said, your company wants every communication from you stored somewhere, ready for analysis. I don't think an unfiltered data acquisition is good, my interpretation and decision making is also part of my work. Also meetings may share some personal details that I would never tell on the record.

Full transparency has a cost, and we cannot afford it.

[wuschel]

Unfortunately you can't record meetings in many jurisdictions, including court sessions. Hence we have to rely - for worse, or perhaps even for better - on human driven note taking.

[poisonfountain]

You're downplaying the AI lobby here. They're eating down copyright laws, something that seemed impossible just a couple of years ago. Screwing privacy laws is just the next step.

Also, we are seeing a cultural shift around that as well. Now people bring "AI notetakers" to Zoom calls without even asking for your permission. People are already acting like privacy laws don't exist anymore, it's going to be even easier for the AI lobby to take it down now. Just like piracy normalized copyright infringement, opening the path to the current rulings around "fair training".

[Yokohiii]

Such invasive practices are pretty disgusting. But I don't think it will be pervasive. Once it spreads, AI vendors and abusive companies will be hold accountable. There is also an obvious conflict, the surveillance will likely be very selective. Programmers have to record everything, while middle managers have a choice to sign off everything. Senior management will of course do whatever but have full insight on the data. This will create even more backlash. Of course the social culture will turn stone cold and hostile over night with such installments.

[Yokohiii]

thanks for the downvote anon. its an convenient conversation.

[poisonfountain]

I disagree but it wasn't me who downvoted, just so you know.

[wood_spirit]

I have found Claude et al good at quickly implementing the algorithm I have in mind effectively, as long as I ask lots of control questions and check code. They aren’t good at inventing non-mainstream algorithms though and often slip staggeringly short term shortcuts in though. They are still a tool and not yet the craftsman who wields tools effectively. This will steadily change, and the corners where the obscure algorithm wins will erode further too.

[cyanydeez]

I'd say it's a malefactor of:

1. Amazing, you just tweaked 1% efficiency

2. You idiot, you just spent an hour trying to trouble shoot a hallucinated api.

On average, it's really hard to tell which ones going to win here.

[dakolli]

Its not hard to tell at all, just look at how much it costs to run a 10T param model (especially with parallelized agents). Those costs are not worth the occasional slot machine-eque jackpot you get. For an entity like Google it might be worth it, but that's it. They definitely aren't going to let us use these things for cost they are now for much longer.

Imagine going back to 2020 and tell people in 6 years going to be able to spend $200.00 a month and be able to spin up $2mm in GPUs at full throttle to respond to your emails. None of this makes sense.

[ogogmad]

Whenever you solve any hard problem, you start off by finding a complicated solution, which you then scale down to a simpler solution.

LLMs are a "complicated solution" in the sense that they're expensive. Once you know what they're capable of, you can scale them down to something less expensive. There's usually a way.

Also, an important advantage of LLMs over other approaches is that it's easy to improve them by finding better ways of prompting them. Those prompting strategies can then get hard-coded into the models to make them more efficient. Rinse and repeat. Similarly, you can produce curated data to make them better in certain areas like programming or mathematics.

[cyanydeez]

they're not _compplicated_, their complex. And solution implies they're not hallucinating the goat and how to fix it.

[Leynos]

You don't pay for a £200 a month account to respond to your emails, and if you are, I would tell you that you're wasting your money.

[throw310822]

I don't know, I guess it depends from a) how many hours per month you spend answering emails, and b) how much more revenue you could get in that same time. $200 should be reasonably 2/3 hours of work? So that's about the amount of saved time per month to break even on your subscription. It's a steal.

[cyanydeez]

oh, sorry, I'm not running a 10T param. Just local models for me. kk thx.

[snapcaster]

Do you realize you're fighting a strawman or do you actually think this is a compelling argument?

[dinfinity]

> I think the rest of us should rest easy knowing that LLM's can't [...]

What if (when?) (AI-assisted) research moves AI beyond LLMs? Do you think that can't happen?

[kubb]

Not in the next decade. Won't get funded.

[dinfinity]

Private investment in the US has grown from 100 billion in 2024 to almost 300 billion USD in 2025 [0]. Add public investments worldwide and private investments in at least China and Europe.

I'm pretty sure money is not going to be the blocker.

[0] https://hai.stanford.edu/ai-index/2026-ai-index-report

[kubb]

The money will go to LLMs.

[dgellow]

Why not both? You don’t need 1trillion allocated before you have a proof of concept to demonstrate your non-LLM model, and once you have a PoC you will definitely have the larger investors interested

[kubb]

You will need 100s of billions to make a viable POC.

[drysine]

Advanced Machine Intelligence (AMI), a new Paris-based startup cofounded by Meta’s former chief AI scientist Yann LeCun, announced Monday it has raised more than $1 billion to develop AI world models.

LeCun argues that most human reasoning is grounded in the physical world, not language, and that AI world models are necessary to develop true human-level intelligence. “The idea that you’re going to extend the capabilities of LLMs [large language models] to the point that they’re going to have human-level intelligence is complete nonsense,” he said. [0]

[0] https://www.wired.com/story/yann-lecun-raises-dollar1-billio...

[holoduke]

Why on earth would you start your ai startup in Paris? Of all places in western Europe it's one of the hardest to find, attract and keep talented people. The wages are super low, housing is high and language is an issue.

[margorczynski]

Probably because LeCun is from there. But top AI talent needs to be paid top cash and the taxes there are brutal for high earners especially.

[kubb]

Now check how much OpenAI got in their last funding round, and you have your answer.

[drysine]

1B is what Microsoft invested in Open AI in 2019[0]. That was enough to get the ball rolling.

[0] https://en.wikipedia.org/wiki/OpenAI#Creation_of_for-profit_...

[DennisP]

I don't think it's valid to draw broad conclusions from the funding of a new company vs. an industry leader. If AMI builds something that looks impressive considering the funding they got, then they'll get plenty more in the next round.

[dinfinity]

He must be trolling.

AI is hands down the most researched topic in CS departments. Of the 10 largest companies (by market cap), only 3 aren't balls-deep in AI R&D. The fastest growing (private or public) companies by revenue are also almost all companies focused primarily on AI (Anthropic, OpenAI, xAI, Scale AI, Nvidia).

And the money isn't even the most important part. It's all about mindshare and collective research time. The architectural concepts can be researched and developed on top of open models, so even individual relatively poor researchers unaffiliated to anything can make breakthroughs.

Even the computing required for the legendary "Attention is all you need" paper could probably be recreated on con-/prosumer hardware in a month's time.

[ActorNightly]

I mean, Google already has Mu Zero, which Im willing to bet has evolved quite a bit in private because if anything is going to get us closer to actual AI its that.

Realistically, one can build a AI capable of reasoning (i.e recurrent loops with branches) using very basic models that fit on a 3090, with multi agent configuration along the lines https://github.com/gastownhall/gastown. Nobody has done it yet because we don't know what the number of agents is required and what the prompts for those look like.

The fundamental philosophical problem is if that configuration is possible to arrive at using training, or do ai agents have to go through equivalent "evolution epocs" to be able to do all that in a simulated environment. Because in the case of those prompts and models, they have to be information agnostic.

[vonneumannstan]

>I think the rest of us should rest easy knowing that LLM's can't (and maybe were never meant to) tackle the tacit-knowledge-filled, human-system-centric, ambiguously-defined-problem-space jobs most mortals work.

A Statement all but guaranteed to look incredibly short sighted by 2030.

[01100011]

The past few years has seen a great rise in casuals reminding us of AIs limitations only to be proven wrong in 6 months. I don't think we're close to AGI, but in 2 years I've gone from AI doubter to AI convert. It's not perfect, but I don't need it to be.

The real question to me is if the system can pay for itself. Economics are racing against efficiency gains and it's anyone's guess which wins.

[byzantinegene]

what are those limitations we're talking about? seems most of those the original limitations that people complained about were resolved through workarounds like tools and skills which are more software-engineering than llm advancement.

[vonneumannstan]

Long-term planning, context poisoning, task length reliability, etc. Improving all the time but certainly still constraints.

[nerdsniper]

The biggest limitation I see right now is weak “theory of mind”. It’s why even though AI can generate very decent exposition, it sucks at generating narrative. This also reflects in weak performance at humor, art, and even shows up in exposition (resulting in reactions akin to “cool story bro, but why should I care?”)

It’s why people can identify AI writing even if it doesn’t contain any LLMisms. AI’s generate text that almost looks like a human wrote it, but that no human would ever actually write - when we try to imagine what kind of person would have wrote this, we draw a blank - no one we’ve ever met would have written it like that - not even any archetype we’ve ever built an internal model for.

[DoctorOetker]

"... tackle the tacit-knowledge-filled, human-system-centric, ambiguously-defined-problem-space jobs most mortals work."

sounds like jobs involving legalese, politics, corruption and more generally involving pretending you don't understand something for which your income depends on not openly understanding something...