You’re welcome. I think calling it the output of an ‘AI model’ triggers thoughts of the current generative image models, i.e. entirely fictional which is not accurate, but it is important to recognise the difference between an image and a photo.
I also by no means want to downplay the achievement that the image represents, it’s an amazing result and deserves the praise. Defending criticism and confirming conclusions will always be vital parts of the scientific method.
Most of what you said is correct but there is a final step you are missing, the image is not entirely constructed from raw data. The interferometry data is sparse and the ‘gaps’ are filled with mathematical solutions from theoretical models, and using statistical models trained on simulation data.
Paper: https://arxiv.org/pdf/2408.10322
We recently developed PRIMO (Principal-component Interferometric Modeling; Medeiros et al. 2023a) for in- terferometric image reconstruction and used it to obtain a high-fidelity image of the M87 black hole from the 2017 EHT data (Medeiros et al. 2023b). In this approach, we decompose the image into a set of eigenimages, which the algorithm “learned” using a very large suite of black- hole images obtained from general relativistic magneto- hydrodynamic (GRMHD) simulations
Its not hard to find that there are legitimate academic criticism of this ‘photo’. For example here. The comparison you made is not correct, more like I gave a blurry photo to an AI trained on paintings of Donald Trump and asked it to make an image of him. Even if the original image was not of Trump, the chances are the output will be because that’s all the model was trained on.
This is the trouble with using this as ‘proof’ that the. Theory and the simulations are correct, because while that is still likely, there is a feedback loop causing confirmation bias here, especially when people refer to this image as a ‘photo’.
This is 100% the answer, the only solution to the model decay from LLM outputs overwhelming the web is to start collecting data IRL.
This is also why companies like OpenAI are desperately investing in ‘AI wearables’ that no-one wants. They have to get the unpolluted data from somewhere, and recording real conversations will at least mostly have come from actual humans instead of AI.
Alternatively, following their logic, keep the number of people and achieve massively higher productivity. But they don’t want that, they want to reduce the number of people having opinions and diluting the share pool, because its not about productivity, its about exerting control.
You’re seeing insufferable people who also happen to use Linux, as insufferable people also like to be early adopters so they can say they are different and therefore better.
There are plenty of super helpful people in the community, and Linux is well past the early adopter phase. The transition from Windows is smoother than it has ever been.
I disagree, I think the removal of ads is often painted as a benefit that had inherent value. Look at YouTube premium or Prime video. Both haven’t actually improved their offering, just made it worse by introducing ads and insisting users that don’t want to see ads have to pay for the privilege of not being advertised to.
This means the total price adds up to higher than 100% of the product value, because it’s a ‘premium’ version that comes without advertisement inconvenience.
Paying for the product and paying to not be inconvenienced by ads have become separate things. The first is standard business, the second is extortion.
Feel free to contribute any features you feel it is missing
This 100%, too many musk fanboys don’t understand how much he is actually hindering innovation.
Even at SpaceX they make more progress when he isn’t involved.
Owning that much of a company that is valued that highly is still damaging to society, even if it isn’t liquid cash. Even putting aside their ability to take out loans with the shares as collateral, if the company is really worth that much it should be owned by a larger number of people with each taking a reasonably sized share to ensure that decisions are not made selfishly.
Taxing unrealised gains also hurt working class people dabbling in the stock market to try and improve their circumstance. IMO once you reach a net worth of $1B you get a pat on the back that you won captalism, and a 200% tax rate on anything beyond that to force you to give it up. No one person should own and control so much of a company if it truely has so much value, divide it among those that created the value i.e. the employees.
You seem to be in the camp of believing the hype. See this write up of an apple paper detailing how adding simple statements that should not impact the answer to the question severely disrupts many of the top model’s abilities.
In Bloom’s taxonomy of the 6 stages of higher level thinking I would say they enter the second stage of ‘understanding’ only in a small number of contexts, but we give them so much credit because as a society our supposed intelligence tests for people have always been more like memory tests.
This is now top of my list
Definitely had issues on first release, but a lot has improved since then without getting much coverage. Btw I wouldn’t say that x86 has ‘caught up’ especially if your metric is power efficiency, not just raw power. Until we see a realistic RISC-V offering arm will likely remain king in that space.
They were humans pretending to be robots, these are actually robots that are being remotely controlled, rather than being fully autonomous.
Pretty sure you already can’t block him specifically, I’m not on Xitter but have heard from several people this is the case.
This is the exact metaphor I’ve been using when talking to people about the issue. Did we both get it from somewhere I can’t remember, or is it just perfect?
Binary Large OBject
Basically any binary file, often objected to in open source repos because of the lack of source and ‘openness’. See also the recent xz backdoor.
It also doesn’t run off battery power, hardly apples to apples.
Thanks, being a software engineer and working in interferometry I was familiar with some of the details - enough to want to jump in when you were getting downvoted - but I will admit I only found and read the actual paper for the first time because of this thread, as I wanted to be sure on the facts!