To add to this insight, there are many recent publications showing the dramatic improvements of adding another modality like vision to language models.
While this is my conjecture that is loosely supported by existing research, I personally believe that multimodality is the secret to understanding human intelligence.
I am an LLM researcher at MIT, and hopefully this will help.
As others have answered, LLMs have only learned the ability to autocomplete given some input, known as the prompt. Functionally, the model is strictly predicting the probability of the next word+, called tokens, with some randomness injected so the output isn’t exactly the same for any given prompt.
The probability of the next word comes from what was in the model’s training data, in combination with a very complex mathematical method to compute the impact of all previous words with every other previous word and with the new predicted word, called self-attention, but you can think of this like a computed relatedness factor.
This relatedness factor is very computationally expensive and grows exponentially, so models are limited by how many previous words can be used to compute relatedness. This limitation is called the Context Window. The recent breakthroughs in LLMs come from the use of very large context windows to learn the relationships of as many words as possible.
This process of predicting the next word is repeated iteratively until a special stop token is generated, which tells the model go stop generating more words. So literally, the models builds entire responses one word at a time from left to right.
Because all future words are predicated on the previously stated words in either the prompt or subsequent generated words, it becomes impossible to apply even the most basic logical concepts, unless all the components required are present in the prompt or have somehow serendipitously been stated by the model in its generated response.
This is also why LLMs tend to work better when you ask them to work out all the steps of a problem instead of jumping to a conclusion, and why the best models tend to rely on extremely verbose answers to give you the simple piece of information you were looking for.
From this fundamental understanding, hopefully you can now reason the LLM limitations in factual understanding as well. For instance, if a given fact was never mentioned in the training data, or an answer simply doesn’t exist, the model will make it up, inferring the next most likely word to create a plausible sounding statement. Essentially, the model has been faking language understanding so much, that even when the model has no factual basis for an answer, it can easily trick a unwitting human into believing the answer to be correct.
—-
+more specifically these words are tokens which usually contain some smaller part of a word. For instance, understand and able would be represented as two tokens that when put together would become the word understandable.
I hate that I am defending Israel when I say this because what is occurring in Gaza is tragic, but a lot of people are confusing “Genocide” for perceived “War Crimes” as defined by international law and also confusing “Hamas” for “Palestine” or the “Palestinian Authority”.
Hamas is terrorist government (similar in nature to the Taliban) that receives a lot of external funding from countries that actively wish to see the death of Israel and all Jews, making Hamas the chief perpetrators of Genocide in this conflict despite how ineffective they have been in their goals.
Israel was attacked by this terrorist government, and is now defending itself with the expressed war goal of destroying Hamas. While Israel has had a tenuous relationship with the Palestinian people (namely the government’s active efforts to limit the Palestinian Authority and drag their feet on grant the PA more autonomy and their own state which is deplorable and inexcusable), they do not and have not wished to kill an entire culture of people.
Complicating matters, Hamas commonly employs warfare techniques that go against the Geneva Convention like placing government and military headquarters in basements of protected buildings like Hospitals and places of worship. The moment they do that, and abuse those international recognized sanctuaries, they become legitimate military targets leading to the tragic deaths of unwitting civilians.
People can object to the war on the grounds that war is tragic and results in many civilian casualties, but to make meritless claims is detrimental to both international institutions and to the definition of a Genocide. South Africa calls what Israel is doing a genocide, but also explicitly looks the other way with Ukraine and continues to forge close ties with Putin? (For the record, Russia’s actions in Ukraine are also not considered genocide under it’s strict international definition, but they have been found guilty of war crimes).
Israel has an internationally recognized right to defend itself, and it is doing that by dismantling Hamas through force. The Palestinian people are unfortunately caught in the crossfire. With that said, Israel’s methods to this end are not above criticism, and they have faced pressure from the US and Biden to limit civilian casualties wherever possible, and use ground forces to directly attack Hamas rather than relying on airstrikes that have resulted in many innocent deaths.
For those reading who think all war is bad, I’ll leave you with this quote from John Stuart Mills:
War is an ugly thing, but not the ugliest of things: the decayed and degraded state of moral and patriotic feeling which thinks that nothing is worth a war, is much worse. When a people are used as mere human instruments for firing cannon or thrusting bayonets, in the service and for the selfish purposes of a master, such war degrades a people. A war to protect other human beings against tyrannical injustice; a war to give victory to their own ideas of right and good, and which is their own war, carried on for an honest purpose by their free choice, — is often the means of their regeneration. A man who has nothing which he is willing to fight for, nothing which he cares more about than he does about his personal safety, is a miserable creature who has no chance of being free, unless made and kept so by the exertions of better men than himself. As long as justice and injustice have not terminated their ever-renewing fight for ascendancy in the affairs of mankind, human beings must be willing, when need is, to do battle for the one against the other.
Not quite, this was made with a ControlNet. A hybrid image wouldn’t work as well as this does. But the underlying visual phenomena is the same.
This is done by combining a Diffusion model with ControlNet interface. As long as you have a decently modern Nvidia GPU and familiarity with Python and Pytorch it’s relatively simple to create your own model.
The ControlNet paper is here: https://arxiv.org/pdf/2302.05543.pdf
I implemented this paper back in March. It’s as simple as it is brilliant. By using methods originally intended to adapt large pre-trained language models to a specific application, the author’s created a new model architecture that can better control the output of a diffusion model.
I wholeheartedly agree with the purpose of this community and what it advocates for, but I wanted to add some rebuttals to your points.
You mention you have the privilege of not using your car. In car centric parts of the world, which is anything that isn’t a big city, this is a privilege. My family that lives in a town of 50,000 people in Germany still need to use their cars every day! There is only a bus system to get them around. Each working age adult uses their car every day! Including the one who lives in Köln and literally walks across the street to his office because car transportation is far more time efficient than transit.
See anecdote above. I live in Boston, which is supposedly extremely transit friendly and the T and commuter rail, while remarkably extensive, are abysmal. I rode it for two months until I finally gave up and got a car. I live in a house 0.5 miles from the commuter rail station and it’s the cheapest around at $750k, I should be able to have reliable transport to MIT/Kendall.
Your issues with driving, being honked at, being annoyed with lack of right of way, all seem to come from inexperience driving in the city. On roads with speed limits of 30 mph, there is no right of way, it’s just about whoever goes first. After some time, you learn what to expect from locals and adapt to their style. But I understand if driving in a big city makes people uncomfortable. There is a lot going on, and a lot to pay attention to.
Schedule. My god is our transit schedule awful. Commuter rail only once every hour. It’s either 5 minutes early or 20 minutes late regularly. So it’s completely unreliable. The Red line is now slow as fuck. Crawling at 10 mph in most areas now. It’s faster to ride your bike between stations, and get stopped by every traffic light than it is to ride the train. And now the red line only has service every 20-30 minutes!
I loved visiting London. We even got a rental to see things like Stonehenge or Brighton, but I never felt the need to use the car much within the city. While I thoroughly enjoyed driving through parts of London like tiny bridges that had inches of clearance on either side of the vehicle, or massive roundabouts near Victoria station, I never needed to do that for local journeys.
It takes 30 minutes for me to drive to work, but 50-90 minutes to go by public transit and I literally live in a massive transit corridor with service from my house directly to work. It’s absolutely absurd. Essentially, transit only broadly works in US if you live in NYC. It’s too sparse in SF to be used widely. Too sparse in DC. Chicago is 10 times worse with it’s urban sprawl. And unreliable as fuck in Boston. Boston doesn’t even have a reasonable train to the goddamned airport (yes I know about the blue line, but it’s still a 15 min bus from the blue line station, and you can only transfer to the blue line from the green line).
This is why people drive. Because for the vast majority off us, even those in Europe, there is no better alternative. If transit was so much cheaper, then why doesn’t every village of 10,000 people in Europe have a tram? There are simply too many places people want to go, and only extreme density can make transit cost effective.
The same argument could be said for an apartment building too. We need to collectively realize that Single Family Houses are a luxury that most of us will never see in our lifetimes. Our grandparents were able to enjoy them at low prices because the US had half the population it does today.
Restrictive building codes that only permit building SFH is the cause of our housing shortage and not short term rentals that consist of 0.2%-1% of all dwellings.
Many leading economist argue for land value tax only as a way to incentivize the most efficient use for our most valuable resource. If land tax was used instead of property tax, a multi-acre plot in a dense urban would be taxed just as much a multi-story apartment building that takes up the same amount of space.
See the Strongtowns article on the subject. https://www.strongtowns.org/journal/2019/3/8/if-the-land-tax-is-such-a-good-idea-why-isnt-it-being-implemented
At it’s core, this is the root cause of the housing crisis. We do not have enough supply. The amount of Airbnb’s that exist is extremely miniscule and the targeting of Airbnbs is an intentional distraction tactic.
Depending on the source, 1% to 0.2% of all dwellings are listed for short-term rental in the US. That’s crazy small and has very little impact on housing prices overall.
The fact of the matter is that Single Family Homes are an incredible luxury that our parents and grandparents were able to enjoy when the country had half as many people as it does now. It is no longer sustainable to expect a SFH in the US, and the American public continuing to cling to that dream and restrictive zoning practices are really what is driving up prices.
If you want an affordable house you will need to move to a rural area where land and labor are cheap. If you want to live near any reasonably sized city, you better be upper middle class to even think about buying a SFH.
We used to run an Airbnb out of the spare rooms in our house. It was very cheaply priced, and we were always booked out for months. Super host status and everything. It was clear most people just look at the price and never the description or rules. We rented two bedrooms with a shared bathroom, and the amount of complaints we received because they had to share a bathroom with someone else was obnoxious.
We closed up shop during the pandemic and just used those rooms as guest rooms instead. In hindsight it wasn’t worth the hassle of dealing with self-centered people who expect an experience superior to that hotels at a quarter the price. We also had some fantastic guests that we loved having stay with us, but the few bad experiences dramatically overshadowed all the good decent people.
Airbnb’s are so shitty today because their customers are just as equally shitty on aggregate.
Cab prices tend to be more consistent across all times of day and location. In Boston an Uber 10 miles west from the airport can cost upwards of $120 depending on the time. A cab would be $60.
The same Uber to the airport is typically much cheaper always at $40.
Just so you know, you can edit titles on Lemmy!
I envy you so much right now! I wish I could forget it all and watch him again. 10 years of content you have the opportunity to go through!
As long as we can all agree NOT Johny Harris. The guy tells an absolutely amazing story, but it is rarely factual. I believe he thinks he is correct, but he lacks a background in most things he touches, and consistently gets in wrong. Pick a subject you know a lot about and watch one of his videos to see what I mean.
For me, it was his video on Space Junk. There are also many other response videos to him. One I think is from an econ professor on Harris’ recent 4 econ videos.
As long as we can all agree NOT Johny Haris. The guy tells an absolutely amazing story, but it is rarely factual. I believe he thinks he is correct, but he lacks a background in most things he touches, and consistently gets in wrong. Pick a subject you know a lot about and watch one of his videos to see what I mean.
For me, it was his video on Space Junk. There are also many other response videos to him. One I think is from an econ professor on Harris’ recent 4 econ videos.
Culturally, 18% is the absolute minimum nowadays. An average tip is north of 20%.
I typically do 20% and round to the next dollar. So if the meal was $56.14, I would calculate $5.60 as approximately 10%, double it to $11.20 and then add 66 cents for a tip of $11.86 do the final total is $68.00.
The servers only enter the total line in the system, so this makes it easier for them.
This is truly a terrible accident. Given the flight tracking data and the cold, winter weather at the time, structural icing is likely to have caused the crash.
Ice will increase an aircraft’s stall speed, and especially when an aircraft is flown with autopilot on in icing conditions, the autopilot pitch trim can end up being set to the limits of the aircraft without the pilots ever knowing.
Eventually the icing situation becomes so severe that the stall speed of the ice-laden wing and elevator exceeds the current cruising speed and results in a aerodynamic stall, which if not immediately corrected with the right control inputs will develop into a spin.
The spin shown in several videos is a terrifying flat spin. Flat spins develop from normal spins after just a few rotations. It’s very sad and unfortunate that we can hear that both engines are giving power while the plane is in a flat spin towards the ground. The first thing to do when a spin is encountered is to eliminate all sources of power as this will aggravate a spin into a flat spin.
Once a flat spin is encountered, recovery from that condition is not guaranteed, especially in multi-engine aircraft where the outboard engines create a lot of rotational inertia.