90% confidence level, Hottest weather in recorded history. Extreme weather in many places not seen before. No one is prepared for this commonplace outlier, Globally common, locally outlier. Huge insurance payout but most not fullfilled.
90% CL: Top search engine results completely inundated by LLM text, we none the wiser. SEO solved. Search engine switches to new algorithm based on LLM.
90% CL: AI with tool use that has access to physics simulator and CAD software will automatically advances efficent engine design.
70% CL: Million dollar prompt run on million dollar math prize. Millennium Prize Problems
40% CL: AI sucessfully held back invasive species encrochment with lessons from AI war machines
30% CL: Cheap AI nose sniffs out many common diseases. Automatic data analysis of Gas Chromatography and spectroscopy readings on patients' VOC(volatile organic compounds) produces fast and cheap diagnosis.
10% CL: Not next year, Cost of renewable electricity in remote area becomes cheap enough to pull oxygen from air or electrosis from water to be sold as commodity. Or other processes to package solar energy for trade.
I read a piece about how abundant solar in rural Afghanistan strengthened the Taliban as poppy and therefore opium production skyrocketed because they were able to install electric wells.
Here in W.Australia we've been using solar powered signal repeaters along rail lines and mesa tops since the mid 1970's in the Pilbara mining industry.
Solar powered bores for farms in the wheatbelt and cattle country has been a regular feature for over a decade now.
The panels and batteries are better than ever now, but they've been Total Cost break even better than fossil fuel alternatives in remote Australia for a very long time (esp. factoring in need to travel to a bore to refuel, etc. Vs "it just works" (until it doesn't, but all solutions require upkeep)).
> Cost of renewable electricity in remote area becomes cheap enough to pull oxygen from air or electrosis from water to be sold as commodity
Oxygen is already kind of a commodity. Did you mean hydrogen? Anyway, this isn't going to happen, look at retail power prices.
> Top search engine results completely inundated by LLM text, we none the wiser. SEO solved. Search engine switches to new algorithm based on LLM.
I'm not sure in what sense SEO is "solved" here, but this basically puts the Internet into a nasty local minimum where it's kind of good enough for information you don't care about, but will feed you wrong information often enough to cause problems, while having stripped all the traditional reliability signals. Possibly also kills what's left of fact-based journalism.
> 90% CL: Top search engine results completely inundated by LLM text
Are you sure you meant this as a prediction for 2025? It has been months, if not years, since search engines have been inundated by LLM articles.
The pursuit of advertising dollars has even made many long-tail queries worthless. Results often are a pile of AI-generated spam sprinkled with display ads and biased affiliate links. Ad blockers can deal with some of that, but there remains the problem of not finding appropriate results to many search queries.
This has been the sad state of the web for quite some time, and it’s just going to worsen.
"Solving" is a bad choice of word. SEO does not have a solved state.
Digital design of engines can already happen now. AI chip design is already a thing, I think they help lay out more efficient trace lines. Since digital design and simulation can all happen digially. The whole cycle is digital. A system like Claude's computer use can take control and produce 100s or thousands of small modification of existing design to be simulated and if it has testable performance critiera. I think it can be done now. Maybe some finetuning the model for specific UI control.
Disease can change body odor, and dogs have been trained to "sniff out disease" [0][1] (i.e. detect via smell) for years. Various researchers have been working on robot noses that would be able to do the same.[2] Presumably a widespread and/or less expensive version of this "robot nose" is what doku means by "AI nose sniffs out many common diseases".
lol, I do mean a sniff with a nose. The technology I'm referring to can work by taking air sample. This technology exist, the prediction is that cheap version will be available. My assumption is robots will want a digital nose, for example nanny bots will want to smell smoke and react. An example I read for this technology is that grocery store robot will want to detect spoiled food. It will be mass produced and gets cheaper. But we don't have a robot industry yet, so my confidence level on this prediction is low. The AI part of this is just data analysis of spectral lines.
I think it's roughly: "I would bet 9$ to your $1 that this happens", although some of these are vague enough that you couldn't objectively determine whether the bet materialized.
AI in cyber defence and attack
Quantum resistant cryptography
Death of the password? Passwordless, passkeys, phishing resistant MFA
More suppliers compromised, another major open source project long con
Doing more with less: automation and consolidation
90% CL: Top search engine results completely inundated by LLM text, we none the wiser. SEO solved. Search engine switches to new algorithm based on LLM.
90% CL: AI with tool use that has access to physics simulator and CAD software will automatically advances efficent engine design.
70% CL: Million dollar prompt run on million dollar math prize. Millennium Prize Problems
40% CL: AI sucessfully held back invasive species encrochment with lessons from AI war machines
30% CL: Cheap AI nose sniffs out many common diseases. Automatic data analysis of Gas Chromatography and spectroscopy readings on patients' VOC(volatile organic compounds) produces fast and cheap diagnosis.
10% CL: Not next year, Cost of renewable electricity in remote area becomes cheap enough to pull oxygen from air or electrosis from water to be sold as commodity. Or other processes to package solar energy for trade.