Dumping our apprenticeship programs onto academia is exacly how we got into this mess to begin with. It has historically not been the job of a college to produce junior talent. They teach a best for T shaped individual and setup for more of their pipeline in research should students want to delve deeper
If industry doesn't want to pay for training, they better pay bootcamps to overhaul themselves and teach what they actually need. I don't think universities will bend much more since they have their own bubble on their hands.
I have a few lines of "download subtitles with yt-dlp", "remove the VTT crap", and "shove it into llm with a summarization prompt and/or my question appended", but I mostly use Gemini for that now. (And I use it for basically nothing else, oddly enough. They just have the monopoly on access to YouTube transcripts ;)
Not doubting you, but what possible purpose could anyone have to use LLMs to output HN comments? Hardly exists a lower-stakes environment than here :) But yeah, I guess it wouldn't be the first time I reply to LLM-generated comments...
Ha — fair point. Hacker News comments are about as low-stakes as it gets, at least in terms of real-world consequence. But there are a few reasons someone might still use an LLM for HN-style comments:
Practice or experimentation – Some folks test models by having them participate in “realistic” online discussions to see if they can blend in, reason well, or emulate community tone.
Engagement farming – A few users or bots might automate posting to build karma or drive attention to a linked product or blog.
Time-saving for lurkers – Some people who read HN a lot but don’t like writing might use a model to articulate or polish a thought.
Subtle persuasion / seeding – Companies or advocacy groups occasionally use LLMs to steer sentiment about technologies, frameworks, or policy topics, though HN’s moderation makes that risky.
Just for fun – People like to see if a model can sound “human enough” to survive an HN thread without being called out.
So, yeah — not much at stake, but it’s a good sandbox for observing model behavior in the wild.
Would you say you’ve actually spotted comments that felt generated lately?
Nothing to do with linear, meaningful projections on embedding spaces, and everything to do with efficient maintenance of legacy data reporting systems.
While the critique is valid, that does not offer a path to the solution.
Utilitarism is the ruling moral philosophy, and the only possible countermeasure is externalities but that depends on an effective government which is even more unlikely that asking for ethical behavior to corporations.
That may be widely believed but there are plenty of government institutions that actually function well. Libraries are a good example.
What’s more: the belief in govt “inefficiency” is one of the hardest to overcome factors that makes it hard to build good institutions, leading to a vicious cycle.
I agree, and that’s the case for dismantling as much of the federal government as possible - it is too big to work. Break up Apple, Google, Amazon, Washington DC.
You’re forgetting the corollary - all small organizations are also inefficient, just at different things.
It’s all trade offs - do you want everyone in your country to have a baseline education that can be relied on as a given but perhaps it suffers in its administration and effectiveness? Or are you OK with pockets of your country having tremendous quality of education and others having very poor quality, as an example.
Public utilities and services are the default and work well in the majority of developed countries. This is true for everything from local transport to water distribution. As the joke says "universal healthcare is so difficult to get right that only all developed countries except the US have managed to put it in place".
There are places where: a) weather predictions are unreliable, b) there is scarcity of water. Just making the right decision on at what hour to water is a huge monthly saving of water.
Need is a very strong word. We don't need a lot of we have today.
But as a hobbyist I would prefer to program in an LLM than learn a bunch of algorithms, and sensor readings. It's also very similar to how I would think about it, making it easier to debug.
I think there’s two schools of thought. The models will get so big everyone everywhere will use them for everything and they will make lots of money on api calls. The models will get cheaper and cheaper computationally on inference that implementing them on the edge will cost nothing and so an LLM will be in everything. Then every computational device will have one as long as you pay a license fee to the people who trained them.
In a greenhouse operation with high-valued crops. Automated control technologies in those applications have been around for decades, and AI is competing with today’s sophisticated control technology designed, operated and continually improved by agriculturists with detailed site-specific knowledge of water (quality, availability, etc.), cultivars, markets, disease pressures, etc.. The marginal improvements AI can make in a process of poor data quality and availability, an existing, finely tuned, functioning control system, and facing the vagaries of managing dynamic living systems are…tiny.
The solution for water-constrained operations in the Americas is move to a location with more water, not AI.
For field crops…in the Americas, land and water is too cheap and crop prices are too low to be optimized with AI at the present era. The Americas (10% of world pop) could meet 70% of world food demand if pressed with today’s technologies…40% without breaking a sweat. The Americas are blessed.
Talk to the Saudis, Israel, etc. but, even there, you will lose more production by interfering in the motivations, engagement levels and cultures of working farmers than can be gained by optimizing by any complex opaque technological scheme, AI or no. New cultivars, new chemicals, new machinery even…few problems (but see India for counter examples). Changing millennia of farming practice with expensive, not-locally-maintainable, opaque technology…just no. Great truth learned over the last 70 years of development.
Just as the other comment "have to" is a very strong word. But there are benefits to it: a) adaptability to local weather patterns, b) no access to WiFi in large properties.
I see. I guess it all boils down to how low power you can make this.
Keep in mind that there are other wireless communication systems that are long range and low power that are specifically designed to handle this scenario
If we're paying for reasoning tokens, we should be able to have access to these, no? Seems reasonable enough to allow access, and then we can perhaps use our own streaming summarization models instead of relying on these very generic-sounding ones they're pushing.
> There's ample evidence that thinking is often disassociated from the output.
What kind of work do use LLMs for? For the semi technical “find flaws in my argument” thing, I find it generally better at not making common or expected fallacies or assumptions.
