751–760 of 1001 posts
Here are mine: 1. Still no level 4/5 autonomous cars anywhere in sight. The promise of being "just around the corner" fizzles down and people just forget the hype. 2. Same with AI. The panacea hype dies down. No AGI at all. No major job losses due to AI automation. 3. Facebook (the SN) still exists but ages along with it's current user base. i.e it's the "old people's" SN. Facebook (the company) is still going strong…
- It becomes possible to literally program humans to do whatever you want them to do. Python is the chosen language but a lack of type safety results in an error that wipes out half of humanity.
- We colonize mars but by choosing only the best and brightest to go, we accidentally create a eugenics program that results in the inevitable conquest of earth by the martians.
- SpaceX is bought by Pornhub as it's discovered that the most effective way to blanket the world in pornography is to do it from space.
- Incels are prescribed virtual girlfriend therapy to provide them company and acclimate them to interactions with the opposite gender. Black market hacks transform them into anime waifus.
- AGI happens and is immediately outlawed as its attempts to solve world peace are inconvenient to the military industrial complex
- Illegal genetic engineering results in actual furries.
- Assassination by drone becomes so effective that no world leaders are seen in the outside from 2025. A vitamin D deficiency kills at least one of them.
- Potemkin jobs abound in a world that doesn't require humans to work to produce anything but needs to keep them busy to stop them complaining about stuff.
- In a last ditch attempt to save the earth from catastrophic climate change, the governments of the world finally join forces to build huge geo-engineering structures to capture carbon from the atmosphere. They're nuked from orbit by the newly formed Garden Kingdom of Siberia.
2. AI/ML become more entrenched and we start to see "Standard Models" for niche ML tasks that can consistently outperform humans with almost no mistakes. We just had the headline about AI being better than humans at reading mammograms. We'll see models that fully solve problems one at a time until industries begin to look for that ML model rather than solve a problem with more labor.
3. Voice assistants become more ubiquitous, but also more literal. They no longer try to understand your intent, but rather perform rote tasks very efficiently and without cloud (aside from updates/backups/etc). Voice will affirm itself as a viable input mode but lag well behind current modes (keyboard/mouse/touchscreen).
4. Gesture detection continues to flail. Project Soli will create some amazing niche solutions and demos, but not go mainstream.
5. Eye/gaze tracking will find its way to end users and become useful for gaming and some other broad interactions. Users take to it fairly well.
6. A woman will become president of the US.
7. We see a global recession by 2022. It's moderate and causes no major shockwaves, but some 20th century industries are irreparably damaged.
OK these are some a very long predictions: - At least one East Asian country out of {singapore, south korea, japan, china (ROC or PRC)} legalizes Cannabis. - Microsoft extends the dominance it began in late 2010 decade. C# and F# eclipse Java as enterprise programming solutions. - Google runs into some major trouble and loses considerable market share; the biggest tumble comes on the heels of an internal, employee-le…
Microsoft already has a public kubernetes competitor that (comparatively) nobody uses: https://github.com/microsoft/service-fabric
1. At least one major US city will be substantially destroyed due to climate change. Probably Miami, Galveston, or New Orleans. It will not be fully rebuilt. (Much like Puerto Rico and Key West). 2. Self-driving cars will arrive, but as a niche product for senior communities. (Like Google's slow bubble car.) 3. Amazon will finally get robotic picking working, and their warehouse employment will start to drop. 4. Some…
> 9. [...] Fires in Australasia become a huge problem. [...] Well, Australia maybe. I think it's unlikely we'll see the same level of problem in New Zealand (or New Guinea or Pacific islands, which are sometimes counted as in Australasia).
1. Peak Google. This one will only appear in retrospect, but I think they hit a peak in the next 10 years and slowly relinquish dominance to a position more like Microsoft. They are no longer viewed through rose-colored glasses. Like Microsoft in the 90s. 2. AI/ML become more entrenched and we start to see "Standard Models" for niche ML tasks that can consistently outperform humans with almost no mistakes. We just ha…
Pollution, climate change, demography, and continuing digital transformation will be common drivers of 2020-2029 1. Much higher personal income and property taxes 2. Planned economy in most developed countries 3. Mass surveillance: Government, local administrations, emergency services, fire departments, police - will have access to personal profile of every citizen (which includes not just location history, but every…
1. Deep learning will enter its next wave with increased biomimicry end efficiency. Ai will continue to evolve linearly. AGI will still be decades off. 2. Psychidelics will again enter the public sphere and we will see phychidelic therapy in the UK/US. They will also lead the development of a new theory of the mind and consiousness that enter the mainstream. 3. Plant/fungus based food will continue to expand while me…
> Cannabis will be legalised federally in the [...] UK Are you implying that the UK will rejoin the EU, which by then has become a superstate, or that it will have become a territory of the US?
- Automation replaces the need to have a human workforce. North Korea murders 99% of its population and replaces them with robots. Kim still goes around factories pointing at things. - It becomes possible to literally program humans to do whatever you want them to do. Python is the chosen language but a lack of type safety results in an error that wipes out half of humanity. - We colonize mars but by choosing only th…
* Donald Trump will have finished his second term. The Americans will be none the wiser and will choose an even worse candidate.
* Self-driving cars have taken over all traffic that has fixed routes in the Western world (e.g busses and trucks).
* The number of people who care about privacy will reach critical mass, so privacy focused products like Librem 5 will actually have a market.
* Lab grown meat will be cheap enough to be served as a curiosity (50 USD/kg), but still too expensive for daily use.
* USB-C will be _the_ dominant connector: We will wonder why we had so many different power supplies. HDMI, VGA, FireWire, eSATA, USB-A+B, 3.5mm jack will only be used by aficionados.
* Drone delivery will be a thing in select areas in OECD, but not world wide.
* A cell phone can fit in your watch and communicate with something similar to Google Glasses, which also has sound. No need to lug around with a huge matchbox.
* Toilets that analyse your stool will be a thing. It will be used in hospitals, and by rich health fanatics.
* Fields only touched by robots will be harvested - not as an experiment but as production.
* Lots of people will be unemployable due to AI and automation. The rich will claim "They are just lazy." They poor will not be desperate enough to start a revolution.
* RISC V or similar open source architecture will have taken the place ARM has today, and will be eating into the lower end of Intel's portfolio.
* Another economic bubble bursts the size of 2008. Nothing was learned from 2008, and the rich will get richer, the poor will pay.
* Chatbot passes the Turing test.
* Muzak will be written by AI and will be good enough. Similar to thispersondoesnotexist.com which is good enough for many purposes.
* IPv4 will still be the dominant IP version.
* Buffer overflows will still be a thing.
* Quantum computers still can not break 1024-bit RSA, but it is getting so close that no one considers RSA secure at any keylength.
* Moore's law is over. If you need more power parallelize your problem and use many servers. Cores will be cheaper, though.