Nobody is out seeking to destroy the lives of artists. Companies and people are just pushing the frontier of AI anyway they can, and it just turns out that images and audio are two directions the frontier can be expanded.
> Companies and people are just pushing the frontier of AI ...
in the way that seems most profitable to them.
They may not be 'seeking to destroy the lives of artists', but that's a false dichotomy. The outcome may very well be that they will. They know, and they don't care.
The Chain of Thought in the reasoning models (o3, R1, ...) will actually express some self-doubt and backtrack on ideas. That tells me there's a least some capability for self-doubt in LLMs.
A Poorman's "thinking" hack was to edit the context of the ai reply to where you wanted it to think and truncate it there, and append a carriage return and "Wait..." Then hit generate.
It was expensive because editing context isn't, you have to resend (and it has to re-parse) the entire context.
This was injected into the thinking models, I hope programmatically.
I look at all media organizations skeptically. There are so many ways to distort the truth besides outright lying, and I notice this with the Times -
both in what they choose to cover and their tone when covering.
With that said, the Times is one of the better media orgs. But IMO they should very much not be trusted blindly.
My media diet is a blend of various sources: The Atlantic, The Economist, The Free Press, Reason, Semaphore, Politico, New Statesman, and Axios. Even the Drudge Report sometimes.
I wish I had more right-leaning sources to follow, but I often find their content inflammatory and rage-bait-ey (before anyone complains that liberal media does this too, yes, I agree. HuffPo and similar are cancer.)
This strikes me as curmudgeonly and unnecessarily contrarian.
While it's true that investors, entrepreneurs, corporations, etc. have a vested interest in AI to the tune of trillions of dollars, the impulse to dismiss this as 90% hype (as the author does) is insane.
We're only three years into this, and we have:
- LLMs with grad student-level competency in everything
- Multimodality with complex understanding of photography and technical documents
- Image generators that can generate high-quality photos, in any style, with just a text description
- Song generators that make pretty decent music and video generators that aren't half bad
- Excellent developer tooling & autocomplete; very competent code generation
This is still early and the foundations are still being laid. Imagine where we'll be in 10 years, assuming even a linear growth rate in capabilities.
Think of what the internet is today, and its permanence in everything, and where it was just 30 years ago.
By all means, resist the hype - but don't go so far in the other direction that your head is in the sand.
Why would we assume linear growth in capabilities and not a logarithmic growth rate? It seems time and time again, it gets harder and harder to make progress.
I think back to using Dragon Natural Dictation in 1998, there seemed to be exponential promise and a ton of excitement in my young mind. But the reality was more logarithmic improvements so it is finally pretty good 25 years later.
Combine an exponential growth in investments (that is inherent to our economy) with a logarithmic return in capabilities, and you get a linear increase in capabilities.
Sorry, these claims are just not true. AI generations in these categories are impressive on release, but blatantly generic, recognizable, predictable and boring after I have seen about 100 or so. Also, if you want to put them to use to replace "real work" outside of the ordinary/pretrained, you hit limitations quickly.
The scaling laws of the Internet were clear from the start.
There is an off-putting sort of attitude on BlueSky ("sneering mockery", I guess?). Same attitude was present on Twitter during the pre-Musk era and seems to have migrated over.
Set aside the cleanup for a second: for pure power generation, nuclear is the ultimate clean energy source. Massive amounts of power, extremely cheap when amortized, liked by all political factions (except old-school '70s-era environmentalist liberals), a great source of jobs from blue to white collar, and no pollution or greenhouse gas emissions.
Turning back to nuclear waste: it's a solved problem. At least, very good solutions already exist. For example, see WIPP in New Mexico [0].
As with all energy sources, there is no perfect, one-size-fits-all solution: use a diverse set of energy sources based on what makes sense for the specific locale you're targeting. Nuclear will make sense for some areas (perhaps dense metros), while solar would make sense in others. Geothermal and natural gas as well.
Doesn't seem like that in practice. Nuclear's generally one of the most expensive per-kwh sources of energy on the grid. And nuclear-renewable diversity doesn't favor nuclear either: renewables absolutely destroy any margins nuclear might have, and both want to share the grid with storage or dispatchable fossil fuel generation like gas turbines.
Interesting, looks like I was out of date on that. Wikipedia shows the LCOE of different energy sources [0] and, while nuclear was at parity with natural gas & wind as the cheapest energy circa 2011, it has since become one of the most expensive (~2.5x solar).
Why did this happen? Why has the cost of nuclear doubled in the past decade?
I generate most of the power my house needs right ON my house. I’m perfectly happy sleeping , eating and living there. The rain that runs off the generators waters my vegetable garden.
Yeah if you're into that it's pretty cool, I am too. I don't think nor trust the average joe with maintaining XX kW of batteries next door.
> Do you want to live a few meters from a nuke?
You probably don't know what a nuke is. Anyways, the worst nuclear reactor catastrophe killed ~30 people directly and ~5000 indirectly. Pollution kills ~250k people par year in Europe alone (coal related pollution is ~30k), cars kill about ~4k people in France, per year.
Why do we accept car deaths as an acceptable cost of transportation but not (the extremely few) nuclear related deaths as an acceptable cost of energy production ? Why do we accept 30k deaths per year due to coal pollution but not 5k death, once, because of nuclear ?
It sounds a lot like "dumb apes scared big boom" more than rational reasoning. That or the proverbial boiled frog
Because those who refuse to be exposed to induced risks can prevent most of it, while those who refuse to be exposed to nuclear's induced risks (for them and their children, and their children, and their children... thanks to nuclear waste) are out of practical and realistic options.
Even at 10x you're still a far cry from fossils related deaths, every single year
> Because those who refuse to be exposed to induced risks can prevent most of it
??? In cities the bulk of deaths are cyclists/pedestrians, not even talking about air pollution due to exhaust fumes being produces en masse right in front of people's flats. 1/5 of road deaths in EU are pedestrians
The entire french nuclear waste, since the 70s, would fit in a cube with a side length of 15m... The one that matter: <3% of the total waste, the rest is low radioactivity/short half life waste
The question is not "which amount of fossil fuel, which amount of nuclear?" but about renewables and nuclear.
>> Because those who refuse to be exposed to induced risks can prevent most of it
> ??? In cities the bulk of deaths are cyclists/pedestrians
Being ultra-alert, or avoiding crossing any street or even avoiding any area where motorized vehicles zoom around is possible. Avoiding most of the effects of a major nuclear accident or of some wandering "hot" nuclear waste (during the upcoming 100000 years or so...) is way, way more difficult.
> Coal power plants are literally generating more radioactive ematerial
This is not about preferring coal to nuclear but about preferring renewables to nuclear.
reply