LLMs as a Cognitive Virus

You can finish reading your friend's favorite book or put it down if you don't like it. Neither are you economically compelled to read it by your workplace or forced to build work on top of it that will become difficult or impossible to modify/maintain if you bin the book.

I get that the title sounds a bit sensational, but there really is something to AI adoption beyond book-sharing.

I'm wondering why that is. Maybe they avoided using it because most people think it means something else. (I've certainly had a hard time talking about them, for this reason!)

You are correct. I frequently do this with religions & specific ideologies which seem to root themselves in people the same way that religions do & it certainly isn’t intended to be a charitable framing.

The core of a virus in biology after all is essentially a packet of information which gets injected into host cells in order to replicate the virus in order to spread itself further afield.

In that case the book would be a virus, and your friend a host.

But it generalizes way wider to ideas. Religions are viruses too for example. With very strong viral mechanisms sometimes. Some religions are so much entrenched so that they have become part of the fabric for entire societies or countries, morphing their culture; the same way that sometimes viral RNA can become part of the DNA of the host.

Friends probably are not wide spreading infection points, the way reviews in professional journals are or the idea that successful people have recommended things, but they are especially successful at making the transmitted meme stick, as you will keep getting reinfected from them over time. Unless you break up that dangerous friendship!!

Nah, sometimes you should just come out say it. Those who get offended need to hear it more than anyone else.

OK, but that is just meaningless pseudo philosophical claim. The actual (small) science part of evolutionary biology does not have perspective about ideas being genes. And the evolutionary biology that does is as much pseudoscience as any other pseudoscience.

The framing of something as a disease is itself slightly into the fascist territory.

Like food with no essential vitamins, or carbon monoxide.

We only ever bother learning something because it's interesting, or useful.

Things get muddier when you consider economy. Businesses exist to make money, and money is time.

AI is just another tool for business to manage time.

The 'dependency' and 'lock-in' are business leader issues honestly.

If you care about your craft, your skills, then you'll keep them sharp. You probably already enjoyed programming and it never felt like work.

On the other hand, if programming is just skills to pay the bills, then AI really can help you be more productive.

But then it's up to businesses to look ahead, manage their resources (human resources lol) and keep the business alive.

That means leaning on AI, simply become 'everyone else is doing it' and you have to compete for dollars.

Most people don't care how the sausage is made as long as it's presented well, and tastes great.

Then taking writing as the source of truth… who even knows if the author is the original sayer. They could just write things down they heard other people say - the birth of IP theft! And yet, journalism

I would have taken him for mostly pro LLM.

It's important to make sure it in fact doesn't, otherwise there is nobody to throw a shade on our time capsule having the benefit of hindsight

Shockingly little linear algebra is required to manufacture the weapons that will probably bring about Armageddon. We've had them for damn near a century. It would be hilarious if these latest games with linear algebra were the impetus to make us actually use them at scale.

They certainly would never guess that the text is being generated by people who mostly spend their time sending and receiving text/video of unthinking, brain rotting, bullshit all day and night.

What are you talking about? Those are absolutely issues that are being worked on right here, right now. This framing is obviously false.

The latter is a matter of policy and AI is not needed for that.

I hate this framing. Nickname checks out.

"Word calculator", "linear algebra"... This is just a variation on it's just a stochastic parrot / it's just a next token predictor.

The implication is that because it is "just" this simple-sounding thing, it can't be a complex thing! Except we know empirically that that doesn't hold. What are humans if not "just" next action predictors, "just" a collection of simple atoms, "just" electrical impulses? Yet clearly we are more than the sum of our parts. There no reason to believe dirty water can be more than the sum of its parts but weird sand can't.

There are still people working on medicine, social issues, biology and so on and so forth. People are trying to do these, despite concerted efforts from the ai companies and from the right wings ideologues trying to stop them.

> they didn’t care to connect the world, explore it, and find others.

Then again, people who claimed to be connecting world intentionally harmed real people and even made themselves complicit in genocide. Yes, I am talking about Facebook. And the ones framing themselves as wanna by exploratory are sociopath billionaires who don't care about damage to real people they cause, even glee it it.

> They assembled in the highest courts of the day and involved the most brilliant minds to discuss whether or not linear algebra will bring about Armageddon.

Citation needed. Most brilliant minds do not discuss that. Minds drunk on hype do.

> They shunned any creative or productive endeavor which dared to automate tedium or build on the works of the public domain (lowercase)

Again, I do not know what you are talking about here. We have both visual artists and musical artists. People create and share their creations. Again, despite ai-boosters trying to demotivate them and convince them it is pointless, people are still creative.

So, if anything, I feel like my learning pace has been greatly accelerated. Your point about working on a broader variety of software is really true for me too. I am working on a game that I likely would have never even started had it not been for LLMs. Mind you, I am still doing the vast majority by hand. However, LLMs have been great for little tutorials, setting out a good game plan, etc. It's like having my own little private tutor.

I know enough to not let the LLMs do everything for me. If they do, then I know I will fall into the trap of something breaking, and then I lack the understanding to fix it. I think they should be used to make life easier, but not easy.

Kinda. We also have incredible rates of cardiovascular disease, obesity, and diabetes.

While I do mostly buy that the mind and the body are connected, I don't think being able to bench press 250lbs is a necessary, fundamentally human capability. I do think being able to think is. Literally Rene Descartes cogito ergo sum. If we lose cognition, executive function, analysis, skepticism, etc., I don't think we're human anymore.

The more I use LLMs I find that their output drifts toward being acceptable to other LLMs. I find myself quite often telling the model: "this is not human readable.". Especially Opus and Fable write in such a way that it is difficult to follow. The often put sentences out of context, the context itself is fragmented. Often start with what you would think is end of a story so when you start reading you are like wtf it is about.

If think the issue might be that LLMs read text all at once whereas humans are somewhat linear. So when LLM puts out text it understands, to human it is close to random gibberish that looks like English.

re: 'but a crutch for those who rely on them.' - I was telling a friend about yesterday's experience this morning and was saying that I could easily see populations of people who might become overly reliant on (or completely subsumed by) these new classes of tools being released. They are compelling in a way that they weren't 18 months ago.

Probably surprising for some, but it's possible (and pretty usual) to write at the speed of thought with a pen.

Why? Claude can just do it for you, can't it? If you're using Claude to do these things, why half-ass it?

Yes.

>Because we've become dependent on them?

Because we're burning the planet down with them.

>Should we also facilitate reversibility in our automobile usage?

Yes, because in places we do people enjoy higher qualities of life and happiness. Along with things like health improvement from doing more excersize.

>What about cooking food? We've become entirely dependent on cooking food.

Moreso we've become dependent on eating processed junk over the past 50 years and gotten fat as shit.

So yea, maybe think a little more deeply about your analogies?

>What about cooking food? We've become entirely dependent on cooking food.

Humans have been the dominant species because of our ability to think, not our ability to get from A to B in any particular way/cook food. If we abstract ~the bulk of our thinking to LLMs, don't we lose something as a species which makes us more vulnerable... Or is it always the case during an automation?

Yes?

We're in the midst of figuring out what the tradeoffs are of LLM usage - all the things you list are already mostly known and found worth it. You trade a tiny amount of time and fuel to cook food in exchange for not getting sick. How is that relevant to anything?

The core idea being that each time humanity was going to exceed the carrying capacity of its environment, we invented a new technology which increased the headroom. Obviously this ratchet only goes in one direction. The only way to go backwards is collapse.

But... It's worked so far?

In evolutionary time, we lost the ability to digest raw meat because we offloaded so much of process of the digestion to the cultural technology of cooking.

Do you like the looks of a future where that happens but for thinking?

There will be some growing pains for some, yes. But overall, I think it will ultimately create more jobs than ever. History is full of examples where people thought a new invention leads to catastrophic job loss, then the Earth had more jobs than ever.

We have enough jobs in 2026 to keep most of the 8.3 billion people employed somehow.

The best place to be in a rush over a cliff edge.

Or in other words: the humanities are good.