By now these dates seem like almost the same honestly.
The Singularity - Official Thread
- Powers
- Posts: 1062
- Joined: Fri Apr 07, 2023 7:32 pm
- Location: a.k.a Lurking, Member, Lorem Ipsum, ..., --- and ººº.
Re: The Singularity - Official Thread
different years can be easily distinguished in my opinion
Not really! I at least differentiate years by notable events. Mostly not my personal ones.
In 2015, for example, from what I remember:
• NASA confirmed presence of water on Mars
• New Horizon space probe flew by Pluto and Dawn space probe flew by Ceres, taking photos and sending them back to Earth
• Microsoft released Windows 10, which is the most popular PC OS today
• United States Supreme Court affirmed same-sex marriage (rather controversial move)
• Church of England appointed its first female bishop
• Facebook reached 1 billion users (I don't consider it something overly positive)
• 10 TB HDDs, 4 TB SSDs, 512 GB SD cards and 200 GB microSD cards became available for consumers
• Radeon 390 came out for $329, bringing 8 GB cards with decent performance to homes of normal people
• multiple 18 cores 36 threads Intel Xeon processors came out and were faster in server workloads than 2014 CPUs
• Witcher III, Cities: Skylines, Metal Gear Solid V, Super Mario Maker, Splatoon, Rocket League, Bloodborne, Rise of the Tomb Raider, Xenoblade Chronicles X, Kerbal Space Program, Ori and the Blind Forest and Assassin's Creed Syndicate video games came out
In 2015, for example, from what I remember:
• NASA confirmed presence of water on Mars
• New Horizon space probe flew by Pluto and Dawn space probe flew by Ceres, taking photos and sending them back to Earth
• Microsoft released Windows 10, which is the most popular PC OS today
• United States Supreme Court affirmed same-sex marriage (rather controversial move)
• Church of England appointed its first female bishop
• Facebook reached 1 billion users (I don't consider it something overly positive)
• 10 TB HDDs, 4 TB SSDs, 512 GB SD cards and 200 GB microSD cards became available for consumers
• Radeon 390 came out for $329, bringing 8 GB cards with decent performance to homes of normal people
• multiple 18 cores 36 threads Intel Xeon processors came out and were faster in server workloads than 2014 CPUs
• Witcher III, Cities: Skylines, Metal Gear Solid V, Super Mario Maker, Splatoon, Rocket League, Bloodborne, Rise of the Tomb Raider, Xenoblade Chronicles X, Kerbal Space Program, Ori and the Blind Forest and Assassin's Creed Syndicate video games came out
Global economy doubles in product every 15-20 years. Computer performance at a constant price doubles nowadays every 4 years on average. Livestock-as-food will globally stop being a thing by ~2050 (precision fermentation and more). Human stupidity, pride and depravity are the biggest problems of our world.
-
- Posts: 1941
- Joined: Wed Oct 12, 2022 7:45 am
Re: The Singularity - Official Thread
Polls at the time of Obergefell v. Hodges showed pretty strong support for the measure. The only reason this issue was and is still debated is that a vocal minority remains.
https://news.gallup.com/poll/183272/rec ... riage.aspx
https://news.gallup.com/poll/183272/rec ... riage.aspx
affix to the previous post, for those interested
As for the year 2018, I remember it by:
• Saudi Arabia finally allowed women to drive
• completion of the 100 000 Genomes Project (hopefully allowing for better understanding of what exactly genes do)
• 143 buildings over 200 metres in height were completed in 2018, which was a record
• 527.7 meters high, 427 000 m² of floor area, very nice-looking CITIC Tower (otherwise known as China Zun) - a supertall skyscraper in the Central Business District of Beijing was completed then
• 18 kilometers long Crimea Bridge was completed in Crimea, 2 kilometers long Cao Lãnh Bridge was completed in Vietnam, 1.1 km Luding Yaye Expressway Bridge was completed in China, 1.4 km Sinamalé Bridge was completed in Maldives in 2018
• Apple became the first public company to achieve a market capitalization of $1 trillion (too much imo)
• Bitcoin price fell down from 18K to 4K USD that year, and because of that, GPUs were slowly dropping in price again (although too pricey imo)
• 8-core AMD Ryzen 2700X offered relatively decent performance at a relatively affordable 329 USD MSRP, 6-core 2600X was not far behind
• 14 TB HDDs became available that year
• Xiaomi Redmi 5 phone launched for a very affordable price of 140 or 170 USD depending on the version
• quite a lot video games came out that year, but the most extraordinary thing in gaming that year was Nintendo Labo in my opinion
This makes every year quite distinct. But I was (and I still am) disappointed with developments overall. Too small, too slow, too much bad stuff happening at the same time. Doesn't really look like Singularity to me! Looks like a gradual evolution of everything.
And as for same-sex marriage, I think that marriage is a word reserved for man-woman or woman-man relationship. I don't really think or talk about such topics to be honest.
• Saudi Arabia finally allowed women to drive
• completion of the 100 000 Genomes Project (hopefully allowing for better understanding of what exactly genes do)
• 143 buildings over 200 metres in height were completed in 2018, which was a record
• 527.7 meters high, 427 000 m² of floor area, very nice-looking CITIC Tower (otherwise known as China Zun) - a supertall skyscraper in the Central Business District of Beijing was completed then
• 18 kilometers long Crimea Bridge was completed in Crimea, 2 kilometers long Cao Lãnh Bridge was completed in Vietnam, 1.1 km Luding Yaye Expressway Bridge was completed in China, 1.4 km Sinamalé Bridge was completed in Maldives in 2018
• Apple became the first public company to achieve a market capitalization of $1 trillion (too much imo)
• Bitcoin price fell down from 18K to 4K USD that year, and because of that, GPUs were slowly dropping in price again (although too pricey imo)
• 8-core AMD Ryzen 2700X offered relatively decent performance at a relatively affordable 329 USD MSRP, 6-core 2600X was not far behind
• 14 TB HDDs became available that year
• Xiaomi Redmi 5 phone launched for a very affordable price of 140 or 170 USD depending on the version
• quite a lot video games came out that year, but the most extraordinary thing in gaming that year was Nintendo Labo in my opinion
This makes every year quite distinct. But I was (and I still am) disappointed with developments overall. Too small, too slow, too much bad stuff happening at the same time. Doesn't really look like Singularity to me! Looks like a gradual evolution of everything.
And as for same-sex marriage, I think that marriage is a word reserved for man-woman or woman-man relationship. I don't really think or talk about such topics to be honest.
Global economy doubles in product every 15-20 years. Computer performance at a constant price doubles nowadays every 4 years on average. Livestock-as-food will globally stop being a thing by ~2050 (precision fermentation and more). Human stupidity, pride and depravity are the biggest problems of our world.
Re: The Singularity - Official Thread
OpenAI have basically confirmed that the new noteworthy development towards AGI is coming soon after the election.
A summary of the paywalled article from reddit
https://www.theinformation.com/articles ... elop-orion
"OpenAI is developing a new AI product called Strawberry (previously Q*), which aims to improve reasoning in complex problems such as math and programming, outperforming existing chatbots like ChatGPT, with a possible launch in fall 2024
Strawberry has been demonstrated to national security officials to showcase its reasoning abilities, with its technology potentially aiding future OpenAI agents designed to solve multistep tasks; it can also handle subjective questions, such as product marketing strategies, and solve complex word puzzles like the New York Times' Connections
OpenAI's business is rapidly growing, with monthly revenue tripling to $283 million from LLM and ChatGPT sales, but losses remain high; Strawberry is generating synthetic data for training Orion, OpenAI's next flagship LLM, helping overcome real-world data limitations and reduce model hallucinations
A smaller, distilled version of Strawberry might be integrated into a chatbot to enhance GPT-4 and ChatGPT performance at a lower cost, though OpenAI faces intense competition from Google, Anthropic, and Meta in developing AI reasoning capabilities"
Full article is long and a copy and paste of it may clog up this forum it can be found in the comment section here
A summary of the paywalled article from reddit
https://www.theinformation.com/articles ... elop-orion
"OpenAI is developing a new AI product called Strawberry (previously Q*), which aims to improve reasoning in complex problems such as math and programming, outperforming existing chatbots like ChatGPT, with a possible launch in fall 2024
Strawberry has been demonstrated to national security officials to showcase its reasoning abilities, with its technology potentially aiding future OpenAI agents designed to solve multistep tasks; it can also handle subjective questions, such as product marketing strategies, and solve complex word puzzles like the New York Times' Connections
OpenAI's business is rapidly growing, with monthly revenue tripling to $283 million from LLM and ChatGPT sales, but losses remain high; Strawberry is generating synthetic data for training Orion, OpenAI's next flagship LLM, helping overcome real-world data limitations and reduce model hallucinations
A smaller, distilled version of Strawberry might be integrated into a chatbot to enhance GPT-4 and ChatGPT performance at a lower cost, though OpenAI faces intense competition from Google, Anthropic, and Meta in developing AI reasoning capabilities"
Full article is long and a copy and paste of it may clog up this forum it can be found in the comment section here
-
- Posts: 1941
- Joined: Wed Oct 12, 2022 7:45 am
Re: The Singularity - Official Thread
I hope this means we're on track for AGI by let's say, November. This new Strawberry thingy better be able to power robots to pass the coffee test or something.
- Cyber_Rebel
- Posts: 439
- Joined: Sat Aug 14, 2021 10:59 pm
- Location: New Dystopios
Re: The Singularity - Official Thread
It might be better to view AGI more logically as a series of "steps" or pieces which come together when complete. Q* sounds like the prefrontal cortex within our brains, giving AI more capable abilities like planning, goal setting, and actual learning. Perhaps even understanding true context, if one does not believe our current models are capable of such.
The idea that it's being used to help train the next model is what's most interesting to me if true. That's literally the beginning of the singularity if Orion (I'd like to believe it's an Orion's Arm reference) is a step beyond in its own reasoning capabilities and is utilized to help create Orion 2. Eventually they will not need human involvement in the process, or we'll cease to understand at some point within the architectural framework how this all works.
Open AI waiting till after the election as some sort of governmental agreement may actually be to their benefit, as it helps them develop something with more OOM than if they simply released earlier. I'm not certain about AGI this soon, but within this decade seems likely at least.
What changes first when we actually do have AGI is the real question. Most people are oblivious to the alien entities fast approaching their backyards. If a near future president decides that the best path towards equity is for everyone to have their own AGI assistant, how would they go about securing this? Perhaps the AGI itself might have a better answer? It's ironically similar to the transhumanist argument, where those who have access to the tech become something greater than baseline humans, only a little more subversive in that these aren't actual physical augmentations, and more cognitive extensions.
The idea that it's being used to help train the next model is what's most interesting to me if true. That's literally the beginning of the singularity if Orion (I'd like to believe it's an Orion's Arm reference) is a step beyond in its own reasoning capabilities and is utilized to help create Orion 2. Eventually they will not need human involvement in the process, or we'll cease to understand at some point within the architectural framework how this all works.
Open AI waiting till after the election as some sort of governmental agreement may actually be to their benefit, as it helps them develop something with more OOM than if they simply released earlier. I'm not certain about AGI this soon, but within this decade seems likely at least.
What changes first when we actually do have AGI is the real question. Most people are oblivious to the alien entities fast approaching their backyards. If a near future president decides that the best path towards equity is for everyone to have their own AGI assistant, how would they go about securing this? Perhaps the AGI itself might have a better answer? It's ironically similar to the transhumanist argument, where those who have access to the tech become something greater than baseline humans, only a little more subversive in that these aren't actual physical augmentations, and more cognitive extensions.
Re: The Singularity - Official Thread
This is why it's an imperative that access must be made equally available to all. I'm a transhumanist too, but I am not rich. I don't want a transhumanist future were only the rich have access and are transhumans while I basically become not just lower on the economic ladder but now a lesser species - I want a future were all can be transhuman and equal, including myself. Same with AGI, if this becomes just another tool to put the rich even further above us then it's best we never achieve AGI at all. What we need is to ensure that AGI becomes a great equalizer for all.Cyber_Rebel wrote: ↑Wed Aug 28, 2024 1:37 pm What changes first when we actually do have AGI is the real question. Most people are oblivious to the alien entities fast approaching their backyards. If a near future president decides that the best path towards equity is for everyone to have their own AGI assistant, how would they go about securing this? Perhaps the AGI itself might have a better answer? It's ironically similar to the transhumanist argument, where those who have access to the tech become something greater than baseline humans, only a little more subversive in that these aren't actual physical augmentations, and more cognitive extensions.
- funkervogt
- Posts: 1247
- Joined: Mon May 17, 2021 3:03 pm
Re: The Singularity - Official Thread
We might all be equally dead.What we need is to ensure that AGI becomes a great equalizer for all.
- Powers
- Posts: 1062
- Joined: Fri Apr 07, 2023 7:32 pm
- Location: a.k.a Lurking, Member, Lorem Ipsum, ..., --- and ººº.
Re: The Singularity - Official Thread
Probably.
Singularity is a relative term
My prediction is:
No Singularity in the strictest sense (but in the year 2050 some will say the Singularity has already happened, as their definition of Singularity is different than mine), 3.5% average per year world GDP growth (inflation-adjusted) and on average 19% improvement of overall computer performance per constant energy or work unit (joule, watt) and per constant price, like for example 1000 USD.
After 15 years this adds 67.5% world GDP (after adjusting for inflation) and makes computers 13.6x more energy efficient and probably also 13.6x affordable for most people on the globe. This is approximately in-line with what has happened in the last 15 years (after you average things). People will be more than 67.5% richer in 15 years, as technological deflation will make some things and services better or cheaper so even with the same amount of money you will be able to buy more or better stuff or services. $10,000 humanoid robots counts as $10,000, but is more useful than things which could be bought for $10,000 in the year 2000.
Simple example:
PlayStation 5 (2020) uses 220 watts while being 6.2x faster than PlayStation 4 (2013), which uses 120 watts. This means 3.38x better energy efficiency after 7 years, just in-line with the 19% per year observation. Additionally, if you look at how much better games perform in practice (PSSR and stuff) on the PS5 Pro (2024), they tend to perform exactly 2x better on the Pro, after 4 years and it uses same amount of energy - 220 watts when fully utilized. Price is a weird thing though.
So yes, I do expect exponential growths, especially in everything to do with AI and robotics, but I don't expect true Singularity, just some elements of it.
No Singularity in the strictest sense (but in the year 2050 some will say the Singularity has already happened, as their definition of Singularity is different than mine), 3.5% average per year world GDP growth (inflation-adjusted) and on average 19% improvement of overall computer performance per constant energy or work unit (joule, watt) and per constant price, like for example 1000 USD.
After 15 years this adds 67.5% world GDP (after adjusting for inflation) and makes computers 13.6x more energy efficient and probably also 13.6x affordable for most people on the globe. This is approximately in-line with what has happened in the last 15 years (after you average things). People will be more than 67.5% richer in 15 years, as technological deflation will make some things and services better or cheaper so even with the same amount of money you will be able to buy more or better stuff or services. $10,000 humanoid robots counts as $10,000, but is more useful than things which could be bought for $10,000 in the year 2000.
Simple example:
PlayStation 5 (2020) uses 220 watts while being 6.2x faster than PlayStation 4 (2013), which uses 120 watts. This means 3.38x better energy efficiency after 7 years, just in-line with the 19% per year observation. Additionally, if you look at how much better games perform in practice (PSSR and stuff) on the PS5 Pro (2024), they tend to perform exactly 2x better on the Pro, after 4 years and it uses same amount of energy - 220 watts when fully utilized. Price is a weird thing though.
So yes, I do expect exponential growths, especially in everything to do with AI and robotics, but I don't expect true Singularity, just some elements of it.
Global economy doubles in product every 15-20 years. Computer performance at a constant price doubles nowadays every 4 years on average. Livestock-as-food will globally stop being a thing by ~2050 (precision fermentation and more). Human stupidity, pride and depravity are the biggest problems of our world.
- Powers
- Posts: 1062
- Joined: Fri Apr 07, 2023 7:32 pm
- Location: a.k.a Lurking, Member, Lorem Ipsum, ..., --- and ººº.
Re: The Singularity - Official Thread
I see no Singularity from here.
- Cyber_Rebel
- Posts: 439
- Joined: Sat Aug 14, 2021 10:59 pm
- Location: New Dystopios
Re: The Singularity - Official Thread
Perhaps it's one of those cases where you can't see the ocean while you are within it. Hindsight will give us better context on where it truly began:
-
- Posts: 1941
- Joined: Wed Oct 12, 2022 7:45 am
Re: The Singularity - Official Thread
(here's the post)
heavy doubts about "The Singularity"
Meanwhile, you unfortunately wait 2 years for 10-20% faster CPUs and 20-30% faster GPUs (at the same price or wattage) or for 8% cheaper memory or SSDs. Washing machine with a built-in "AI" doesn't help with anything. It just costs more. It's still the humans who are doing the chores and work, even on computers. From my experience and my acquaintances experience, "AI" is literally almost half the time wrong. What exactly has been getting that much better in the last 10 years? AI-specific hardware? Hardly anyone uses it. Renewables are a bigger change than that. You actually see them on roofs and on grass. No AR glasses, no humanoid robots, no human augmentation, no space travel. Perhaps in the year 2100....
Global economy doubles in product every 15-20 years. Computer performance at a constant price doubles nowadays every 4 years on average. Livestock-as-food will globally stop being a thing by ~2050 (precision fermentation and more). Human stupidity, pride and depravity are the biggest problems of our world.
-
- Posts: 1941
- Joined: Wed Oct 12, 2022 7:45 am
Re: The Singularity - Official Thread
IDK AI doing pretty fine for me. Once again: *sigh* What is your intended use case?
Re: heavy doubts about "The Singularity"
AI still gets things wrong sure - but compare where it is now to just a few years ago, never mind a few years before that. We still haven't miniaturized the technology to make AR glasses useful and practical - but look where VR and AR is on the whole compared to several years ago. Same for a lot of technology - it feels like you're so impatient that it's not here yet and so disappointed that it hasn't been accelerating at the pace you wanted and imagined it should that you've become blind to how much has changed, how fast that change has occurred comparatively, and the actual real world build up of this acceleration trend. It's not going to be this decade, no - it was naive and unrealistic to ever have expected that in the first place. But thinking it'll take another 100 years is going too far in the other extreme. What you want, what you think should be occurring now, is 2030s and 2040s technology. Like AI is breaking so many benchmarks that whether it achieves AGI or not doesn't really matter at all - it'll still be so good and advanced soon that it's going to speed things up so fast in the second half of this decade that it'll be almost impossible to argue that a lot of the stuff we've been talking about for years now won't be reality in the next couple of decades. The biggest issue isn't going to be if the technology's possible, but just scaling up manufacturing and gathering the raw resources needed.Tadasuke wrote: ↑Fri Oct 04, 2024 1:42 pm Meanwhile, you unfortunately wait 2 years for 10-20% faster CPUs and 20-30% faster GPUs (at the same price or wattage) or for 8% cheaper memory or SSDs. Washing machine with a built-in "AI" doesn't help with anything. It just costs more. It's still the humans who are doing the chores and work, even on computers. From my experience and my acquaintances experience, "AI" is literally almost half the time wrong. What exactly has been getting that much better in the last 10 years? AI-specific hardware? Hardly anyone uses it. Renewables are a bigger change than that. You actually see them on roofs and on grass. No AR glasses, no humanoid robots, no human augmentation, no space travel. Perhaps in the year 2100....
It's weird to see you falling so far into despair over the pace of technology when it's so obvious we're on the cusp of major technological change. AR glasses in the 2020s suck - in the 2030s they'll become practical and widely available (whether they prove popular or not is anyone's guess).
As for robots, that's more difficult to predict - there's going to be a lot of pushback from people who are afraid for their jobs due to the whole work-to-survive model we got going...
The Singularity is in a pickle
In my opinion there's no Singularity without next paradigm of computing hardware and even with the next paradigm of computing hardware, Singularity without human augmentation is dystopia or extinction. It wouldn't be awesome for the average unaugmented "Joe".
With the current paradigm, compared to the most cutting-edge stuff, according to my "research", probably (I don't guarantee it's 100% correct):
- CPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- GPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- AI processing units might get 6x faster at the same price and power draw
- memory might get 3x less expensive and 4x more dense than now
- solid state storage might get 3x cheaper and 4x more dense than now
- software might get ~10x more efficient than today
And that's pretty much it for the current paradigm. And if there is the next paradigm, but humans are approximately like today, I see the future as very bleak (for the vast majority of humans at least).
With the current paradigm, compared to the most cutting-edge stuff, according to my "research", probably (I don't guarantee it's 100% correct):
- CPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- GPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- AI processing units might get 6x faster at the same price and power draw
- memory might get 3x less expensive and 4x more dense than now
- solid state storage might get 3x cheaper and 4x more dense than now
- software might get ~10x more efficient than today
And that's pretty much it for the current paradigm. And if there is the next paradigm, but humans are approximately like today, I see the future as very bleak (for the vast majority of humans at least).
Global economy doubles in product every 15-20 years. Computer performance at a constant price doubles nowadays every 4 years on average. Livestock-as-food will globally stop being a thing by ~2050 (precision fermentation and more). Human stupidity, pride and depravity are the biggest problems of our world.
Re: The Singularity is in a pickle
Problem is you're strangely convinced the current paradigm isn't going to change. We're investing and researching heavily into the next paradigm of computing hardware, we're receiving news on the advancements for it all the time here. And I don't know why you think human augmentation isn't going to be a thing when we're seeing remarkable advancements recently in BCI, advanced prosthetics, and even the very first glimmers of useful nanobots.Tadasuke wrote: ↑Mon Oct 14, 2024 12:10 pm In my opinion there's no Singularity without next paradigm of computing hardware and even with the next paradigm of computing hardware, Singularity without human augmentation is dystopia or extinction. It wouldn't be awesome for the average unaugmented "Joe".
With the current paradigm, compared to the most cutting-edge stuff, according to my "research", probably (I don't guarantee it's 100% correct):
- CPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- GPUs might get 4x faster at the same price and power draw (2x single-core, 2x core count)
- AI processing units might get 6x faster at the same price and power draw
- memory might get 3x less expensive and 4x more dense than now
- solid state storage might get 3x cheaper and 4x more dense than now
- software might get ~10x more efficient than today
And that's pretty much it for the current paradigm. And if there is the next paradigm, but humans are approximately like today, I see the future as very bleak (for the vast majority of humans at least).
It's like you've only looked at the technological change of the past decade or so, and feeling so distraught over that have convinced yourself that's the pace it's going to remain for the next ten to twenty years while ignoring all the news and science and experts today. Like you have access to all this news like we do, and yet it seems to make no impact - we've been seeing so many advancements in labs and colleges and companies, been watching as all these things have entered testing, and yet you're blind that those things are coming. Somehow you think the next ten years will be just like the last ten.