frank_nitti

The incentives are there. I know reddit and HN it seems like lots of people are using models/harnesses from scrappy competitors, but IME at the enterprise level I’m in the extreme minority who is even aware of them. Virtually everybody uses Anthropic and OpenAI tools and models, with Copilot getting some scraps because of its forced inclusion in Microsoft/github ecosystem

There is no question about the benefit when things go right; a wary engineer responsibly guiding agents to build orders of magnitude more than was possible before, and always skeptical and critical of the vendor, knowing the vendors’ incentives.

But that’s not the majority, and the people approving absurd token spend are too far out of the weeds to see what’s happening - people burning 6 figures in tokens and still end up with broken systems that keep them over the weekend, but now they hardly understand what they’re troubleshooting (if at all).

This is a dream scenario for the fat-cat vendors, so there is incentive to have CC/Codex write more code, overly-complex designs. When users get upset, they jump to OpenAI to spite Anthropic, or vice versa, but always getting further locked in to the duopoly

Joomla!

I understand what you’re saying now, basically boils down to “might is right”.

It’s not a particularly helpful paradigm for solving predicaments without excess violence and suffering, but the model makes sense insofar as it’s consistent

Perhaps you’re considering a very select subset of the rich for whom you can argue their wealth was entirely earned through value they delivered to society.

Your statement, as written, ignores that much wealth, power, privilege is inherited from ill-gotten gains like savage conquest, slavery, oppression, corruption, etc. and most great wealth is held by people who had extreme advantages from those endowments being born into those circumstances.

In most cases, it’s very hard to argue that society “owes them” anything.

I would say that debt would be much more appropriate for people like Nikola Tesla who wanted his inventions shared with the world at no cost, or the inventors of insulin who sold the patent for $1 to benefit the common good. Not for the telecom and pharma CEOs who become billionaires by extracting the maximum value from exploiting others’ ingenuity!

When billionaires make the decisions about what roads get built, they build roads that help them more efficiently exploit the world around them. This has been observed in many developing countries when their industrialization is driven by private capitalist interests: roads emerge from the plantations to the refineries, to the factories and distribution centers. Roads for the local populations are built to get them to/from these places because capital needs them for labor; roads are not built for them to get to and from their e.g. churches or places of recreation where there is no direct value to be extracted by capital.

When roads are built for the common good, it’s only when government acts on behalf of the working class to compel taxpayers to fund it; it must coerce the billionaires to contribute under the implicit threat of force.

I’d love to believe that humans become generally more collectivist when they amass great wealth - and much credit to those who do - but they are much too few and far between for us to rely this as a society

Interesting, my experience in classes using set theory was the opposite, where ∋ only meant “such that” and ∈ only meant “is a member of”.

Learning programming syntax at the same time made it frustrating to learn that math symbols were less strictly defined and less universal, that it was best to write proofs/derivations/etc in plain English in many cases instead of the neat symbols

In my brief exposure of about 6 months here after ~40 years in Southern CA, it really seems to be the case. I’ve never seen so many people just interacting and enjoying one another’s company for hours on end.

For a decent portion of any given day, nearly every table at every establishment is occupied with people chatting, not browsing nor texting. The local parks are filled with people of all ages playing. Couldn’t help but laugh in disbelief initially

This also happened in the LA area back around 2015, lasting about 36-48 hours - no power and consequently no internet. Out in suburbia, it was the first time many neighbors even met each other, or the first time some neighbors had spoken in person in years.

Standing in our driveways chatting, lending tools or supplies to one another, what used to be very standard suburban life.

It was amazing that we had become so disconnected in only 5 years after smartphones became nearly ubiquitous in that part of the world

Not my comment, but I’d venture to guess they’re referring to the likes of DeepSeek et al, who are/will be able to host their top-tier inference infra more efficiently

I would argue they are also cheating other students in their chosen field, and any future employers who place high value on prestige of applicants’ university, no?

If the only tangible, marketable value of graduating from a prestigious school was the raw knowledge and skill, I would agree with you. But it’s not.

Having worked with people who clearly got preferential treatment in hiring based on their school’s prestige, over more capable applicants from lower-tier schools, I absolutely lose respect for staff at universities who turn a blind eye to cheating.

Interesting to see your observation where I have observed the opposite: posts that share big news about open-weight local models have many upvoted comments arguing local models shouldn’t be taken seriously and promoting the SOTA commercial models as the only viable options for serious developers.

Here and on AI tech subreddits (ones that aren’t specifically about local or FOSS) seem to have this dynamic, to the degree I’ve suspected astroturfing.

So it’s refreshing to see maybe that’s just a coincidence or confirmation bias on my end.

Seems to me the commenter was asking: what observations led us to conclude that original affirmative statement that “the AI did this entirely on its own”.

Given that this is a common technique and not a novel invention, it’s probably present in the training set.

The “surely” reads like it’s referring to the presence of that information in the training set. But your response casts it as saying “surely the AI has not invented something on its own”.

The original question stands IMO, the burden of proof is on whoever is asserting that the AI has invented something on its own, with or without training data that surely already mentions this approach

So perhaps this is just semantics - when we say that “coders have been completely replaced”, to me that means all humans capable of reading/writing code are replaced. In the bridge analogy this is the Licensed Engineer who actually understands and can critically evaluate a system design/implementation in depth.

If the only point being made by “all coders are replaced” is that humans aren’t manually typing the code from their keyboard anymore, I don’t think there’s much interesting to argue there, typing the code was never the hard part.

Who is liable for the runtime behavior of the system, when handling users’ sensitive information?

If the person who is liable for the system behavior cannot read/write code (as “all coders have been replaced”), does Anthropic et al become responsible for damages to end users for systems its tools/models build? I assume not.

How do you reconcile this? We have tools that help engineers design and build bridges, but I still wouldn’t want to drive on an “autonomously-generated bridge may contain errors. Use at own risk” because all human structural engineering experts have been replaced.

After asking this question many times in similar threads, I’ve received no substantial response except that “something” will probably resolve this, maybe AI will figure it out