LLMs aren't AGI and maybe aren't a path to AGI, but step back and look at the way the world is changing. Hard disks were invented by IBM in 1953 and now less than a hundred years later there's an estimated million terabytes a year of hard disks made and sold, and a total sold of Mega, Giga, Tera, Peta, Exa, Zetta ... 1.36 Zettabytes.
In 2000, webcams were barely a thing, audio was often recorded to dictaphone tapes, and now you can find a recorded photo or video of roughly anyone and anything on Earth. Maybe a tenth of all humans, almost any place, animal, insect, or natural event, almost any machine, mechanism, invention, painting, and a large sampling of "indoors" both public and private, almost any festival or event or tradition, and a very large sampling of "people doing things" and people teaching things for all kinds of skills. And tons of measurements of locations, temperatures, movements, weather, experiment results, and so on.
The ability of computers to process information jumped with punched card readers, with electronic computers in the 1950s, again with transistors in the 1970s, semiconductors in the 1980s, commodity computer clusters (Google) in the 1990s, maybe again with multi-core desktops for everyone in the 2000s, with general purpose GPUs in the 2010s, and with faster commodity networking from 10Mbit to 100Gbit and more, and with SATA, then SAS, then RAID, then SSDs.
It's now completely normal to check Google Maps to look at road traffic and how busy stores are (picked up in near realtime from the movement of smartphones around the planet), to do face and object recognition and search in photos, to do realtime face editing/enhancement while recording on a smartphone, to track increasing amounts of exercise and health data from increasing numbers of people, to call and speak to people across the planet and have your voice transcribed automatically to text, to realtime face-swap or face-enhance on a mobile chip, to download gigabytes of compressed Wikipedia onto a laptop and play with it in a weekend in Python just for fun.
"AI" stuff (LLMs, neural networks and other techniques, PyTorch, TensorFlow, cloud GPUs and TPUs), the increase in research money, in companies competing to hire the best researchers, the increase in tutorials and numbers of people around the world wanting to play with it and being able to do that ... do you predict that by 2030, 2035, 2040, 2045, 2050 ... 2100, we'll have manufactured more compute power and storage than has ever been made, several times over, and made it more and more accessible to more people, and nothing will change, nothing interesting or new will have been found deliberately or stumbled upon accidentally, nothing new will have been understood about human brains, biology, or cognition, no new insights or products or modelling or AI techniques developed or become normal, no once-in-a-lifetime geniuses having any flashes of insight?
That's like saying "you're delusional if you think we're affected by The Sun's gravity when it's a hundred million miles away".
A hundred million years ago, every day on Earth was much like every other day and you could count on that. As you sweep forwards in time you cross things like language, cooperation, villages, control of fire, and the before/after effects are distinctly different. The nearer you get to the present, the more of those changes happen and the closer they happen, like ripples on a pond getting closer to the splash point, or like the whispers of gravity turning into a pull and then a crunch. "Singularity" as an area closer to the splash point where models from outside can't make good predictions keeps happening - a million years ago, who would have predicted nations and empires and currency stamped with a human face? Fifty thousand years ago, who could have predicted skyscrapers with human-made train tunnels underground beneath them, or even washing bleached white bedsheets made from cotton grown overseas? Ten thousand years ago, who could have predicted container shipping through the human-made Panama canal? A thousand years ago who could have predicted Bitcoin? Five hundred years ago, who could have predicted electric motors? Three hundred years ago who could have predicted satellite weather mapping of the entire planet or trans-Atlantic undersea dark fibre bundles? Two hundred years ago, who could have predicted genetic engineering? A hundred and fifty years ago, who could have predicted MRI scanners? A hundred years ago, who could have predicted a DoorDash rider following GPS from a satellite using a map downloaded over a cellular data link to a wirelessly charging smartphone the size of a large matchbox bringing a pizza to your house coordinated by an internet-wide app?
In 2000 with Blackberry and Palm Treo and HP Journada and PalmPilot and Windows Phone and TomTom navigation, who was expecting YouTube, Google Maps with satellite photos, Google StreetView, Twitch, Discord, Vine, TikTok, Electron, Amazon Kindle with worldwide free internet book delivery, or the dominance of Python or the ubiquity of bluetooth headphones?
Fifty years ago is 1975, batteries were heavy and weak, cameras were film based, bulbs were incandescent, betamax and VHS and semiconductors were barely a thing - who was predicting micro-electromechanical timing devices, computer controlled LED Christmas lights playing tunes in greetings cards, DJI camera drones affordable to the population, Network Time Protocol synchronising the planet, the normality of video calling from every laptop or smartphone, or online shopping with encrypted credit card transactions hollowing out the highstreets and town centers?
The strange attractor at the end of history might be a long way away, but it's pulling us towards it nonetheless and its ripples go back millions of years in time. It's not like there's (all of history) and then at one point (the singularity where things get weird). Things have been getting weird for thousands and thousands of years in ways that the people before that wouldn't or couldn't have predicted.
Tomy made the Dustbot robot vacuum in 1985, Electrolux made the Trilobite robot vacuum in 1996, and then washing machines, dishwashers, tumble-dryers, microwaves, microwave meals, disposable diapers, fast fashion, and plug-in vacuums, floor steamers, carpet washers, home automation for lights and curtains, central heating instead of coal/wood fires and ash buckets, fridge-freezers and supermarkets (removing the need for canning, pickling, jamming, preserving), takeaways and food delivery, people having 1-2 children instead of 6-12 children. The amount of human labour in housework has plummetted since 1900.
The only solution to people driving is viable alternatives to driving.
Under the previous Conservative government, half of UK bus routes ( ~8,000 ) were cancelled[1]. HS2 high speed train route phase 2 the extensions from Birmingham to Manchester and Leeds - which would move freight as well as people, freeing up space on local train lines for better passenger transport - was cancelled[2]. Phase 1 of it was due to be opened in 2026-2033 timeframe but was bungled now has no planned opening time, and Reform are calling to scrap that, too. Local council budgets were reduced[3] under the austerity measures, including one consequence of 40% less transport spending. The West Coast mainline was sold from VirginRail to Italy's TrenItalia in 2019[4] (Deutsche Bahn, French SNCF and Dutch Nederlandse Spoorwegen own most of the other UK railways) although this government is bringing Rail them back into public ownership.
And Reform are promising to remove bike lanes, and scrap Low Traffic Neighbourhoods to let cars use residental roads as through-roads again[5][6].
The UK doesn't have it as bad as the USA - but that's not for lack of trying to make car the only way to move.
The prompt was "a pelican riding a bicycle"; not prepositions but every verb. Potentially every adverb+verb combination - "a pelican clumsily pushing a bicycle"
I think starting from "put surveillance cameras everywhere, because IF they contribute to a number of arrests that MIGHT end up being a net good" is an uncompelling argument.
How about you give me your bank account details and permission to make free use of all your money, and then IF I contribute more value with your money than you do that MIGHT end up being a net good, so let's try it first and find out later?
"stop reading about the abuses of power, just ignore them" is a really poor position to take.
> I think starting from "put surveillance cameras everywhere, because IF they contribute to a number of arrests that MIGHT end up being a net good" is an uncompelling argument.
I'm not starting from this argument, though. I'm starting from the argument that we shouldn't be inherently opposed to putting up more of we think the benefit outweighs the negatives or risks of abuse
Your level of strawmanning here tells me you're not arguing in good faith
Nobody is "forcing" but billions of dollars on the best psychologists and marketing is doing their hundred year long best to get as close as they can to it.
They (collectively) have thousands of employees, billions of dollars, and tons of things to play with (taste, texture, appearance, mouthfeel, snappiness, salt, sweet, sugar, packaging colours and styles, slogans, marketing, brand associations, shop placement, advertising) all you have is willpower to fight against 4 billion years of evolution telling you that fatty, sugary, energy dense food is important, and a whole pile of biases and weaknesses that you aren't really aware of and they are very aware of and trying to maximally take advantage of.
- Electric motors were invented in 1827 and one of the first things the inventor did was a toy electric car in 1827. The first proper electric car was 1890, and they were popular as golf carts and milk floats through the 1900s. What they were lacking was good enough batteries. - https://en.wikipedia.org/wiki/Electric_vehicle#History
- Autonomous cars: human chauffeurs and taxis (and trains) so we can be moved around without doing the driving, go back to the first cars. We haven't had the technology to build them (and arguably don't and won't have until we get near AGI).
- Data centers in space ... ???? Bueller? Bueller?
- Humanoid robots were seen in Fritz Lang's Metropolis film in 1927, they've always made sense. What doesn't make sense is lying about having built humanoid robots and then having to admit they were being remote controlled, cough Tesla.
In 2000, webcams were barely a thing, audio was often recorded to dictaphone tapes, and now you can find a recorded photo or video of roughly anyone and anything on Earth. Maybe a tenth of all humans, almost any place, animal, insect, or natural event, almost any machine, mechanism, invention, painting, and a large sampling of "indoors" both public and private, almost any festival or event or tradition, and a very large sampling of "people doing things" and people teaching things for all kinds of skills. And tons of measurements of locations, temperatures, movements, weather, experiment results, and so on.
The ability of computers to process information jumped with punched card readers, with electronic computers in the 1950s, again with transistors in the 1970s, semiconductors in the 1980s, commodity computer clusters (Google) in the 1990s, maybe again with multi-core desktops for everyone in the 2000s, with general purpose GPUs in the 2010s, and with faster commodity networking from 10Mbit to 100Gbit and more, and with SATA, then SAS, then RAID, then SSDs.
It's now completely normal to check Google Maps to look at road traffic and how busy stores are (picked up in near realtime from the movement of smartphones around the planet), to do face and object recognition and search in photos, to do realtime face editing/enhancement while recording on a smartphone, to track increasing amounts of exercise and health data from increasing numbers of people, to call and speak to people across the planet and have your voice transcribed automatically to text, to realtime face-swap or face-enhance on a mobile chip, to download gigabytes of compressed Wikipedia onto a laptop and play with it in a weekend in Python just for fun.
"AI" stuff (LLMs, neural networks and other techniques, PyTorch, TensorFlow, cloud GPUs and TPUs), the increase in research money, in companies competing to hire the best researchers, the increase in tutorials and numbers of people around the world wanting to play with it and being able to do that ... do you predict that by 2030, 2035, 2040, 2045, 2050 ... 2100, we'll have manufactured more compute power and storage than has ever been made, several times over, and made it more and more accessible to more people, and nothing will change, nothing interesting or new will have been found deliberately or stumbled upon accidentally, nothing new will have been understood about human brains, biology, or cognition, no new insights or products or modelling or AI techniques developed or become normal, no once-in-a-lifetime geniuses having any flashes of insight?
reply