Kicking the year 2025 off with some predictions. I guess we can return to this in December to see how far they have progressed into fantasy land.
https://www.technologyreview.com/2025/01/03/1109178/10-breakthrough-technologies-2025/
01. Vera C. Rubin Observatory in Chile
02. Generative AI search
03. Small Language Models
04. Cattle burping remedies
05. Robotaxis
06. Cleaner jet fuel
07. Fast-learning robots
08. Long-acting HIV prevention meds
09. Green steel
10. Stem-cell therapies that work
Then they add some potential runner-ups such as Brain-computer interfaces, Methane-detecting satellites, Hyperrealistic deepfakes and Continuous glucose monitors.
https://technologymagazine.com/articles/top-10-trends-of-2025
01. Agentic AI
02. AI governance platforms
03. Disinformation security
04. Postquantum cryptography
05. Ambient invisible intelligence
06. Energy-efficient computing
07. Hybrid computing
08. Spatial computing
09. Polyfunctional robots
10. Neurological enhancement
We are already post-quantum? I wasn't aware that we even had any meaningful utilization of actual working quantum cryptography. Is this the Quantum Leap?
Also I can't help to notice that there seems to be a lot of AI fantasies involved in the predictions for the coming months.
Do you care to make any 2025 predictions of the next big thing?
(Score: 5, Funny) by Tork on Thursday January 16, @12:06AM (1 child)
AI "powered" ad-blockers.
🏳️🌈 Proud Ally 🏳️🌈
(Score: 3, Interesting) by Tork on Thursday January 16, @05:24AM
🏳️🌈 Proud Ally 🏳️🌈
(Score: 2) by donkeyhotay on Thursday January 16, @01:30AM (4 children)
Continuous Glucose Monitors (CGMs) can already be purchased over the counter. And that's the way it should be. It's not like you can overdose on them.
(Score: 3, Touché) by ikanreed on Thursday January 16, @02:24PM
Knowing the American public, I think they'll find a way to prove you wrong.
(Score: 3, Insightful) by VLM on Thursday January 16, @02:42PM (1 child)
For the diet-caused T2 diabetics, pretty much true until they start to take insulin and get into the kidney failure range.
For the T1 folks the "danger" is they're going to F up using it, and then "the computer app told me to take either 0 or 100000 units of insulin so they did" and then the heirs want to sue etc. The idea is to spread the blame and now the PCP or whomever is prescribing it and the pharmacist are now in the blame loop.
If you're just trying a N=1 science experiment of can I personally eat oatmeal without my blood sugar spiking (hint: Probably not for most people) then its pretty safe. Nobody (nobody not on insulin) ever died from not eating too few oreos or not drinking enough corn syrup soda.
(Score: 2) by donkeyhotay on Friday January 17, @12:44AM
I think you misunderstood me. What I meant is that it's not possible to overdose on too many CGMs. Obviously, you can overdose on insulin (as a T1D, I've come close to doing it a couple of tmes).
(Score: 3, Funny) by DannyB on Thursday January 16, @06:20PM
Imagine the continuous glucose monitor challenge! See if you can attach more than 100 of them to your body! The winner will get a box of Tide pods!
Stop asking "How stupid can you be?" Some people apparently take it as a challenge.
(Score: 2) by VLM on Thursday January 16, @03:04PM
There's a fair amount of stuff out there. A lot depends on what you mean by "working" (As if pre-quantum crpyto is "working") and meaningful utilization. The marketing people haven't started pushing it yet so its mostly amateur github repos. Kind of like PGP in the early 90s there's not much commercial interest or use.
https://github.com/PQClean/PQClean [github.com]
One major problem with post-quantum crypto is the definition. People who don't know anything about it, think PQ magically has been proven to be unsolvable. Not so. There's pre-quantum algos that theoretically could be solvable with a quantum computer that doesn't exist but could in the future using algos that have been simulated on a small scale to work. And there's post-quantum algos that do not have a public unclassified quantum algo for that family of math problems. Like lattice problems, multivariate problems, those strange ECC-based problems. Note that AFAIK classic symmetric key protocols like AES have no quantum secret sauce or at least its minimal; AFAIK something like AES with more bits would be quantum-proof as there's no significant quantum speedup algo for those.
Something like UoV looks secure assuming there's no quantum speedup. There was a paper three years ago, about 20 years after UoV was designed, about an interesting hack on UoV that reduced the cost of recovering the key by a factor of up to 2**73. My guess is its cheaper and easier to encourage industry to implement "buggy" crypto than it is to break strong PQ crypto. Hard to say how many more times they'll get away with that strategy. The other issue is solving big multivariate problems seems like something mathematicians would really like to accomplish... so there's pretty intense pressure to discover something that would result in multivariate crypto no longer being post-quantum.
Another interesting point to make is the day of the signature being a tiny little blurb on a large document is likely over in PQ world. UoV sigs are hundreds of K. A rounding error compared to a DVD iso but pretty big compared to a single line IRC chat post.
(Score: 3, Informative) by VLM on Thursday January 16, @04:21PM (1 child)
This is one of those multi-layer onions.
Germany makes 1/3 of the steel in the EU and that takes coke which takes coke ovens and coal and they're aggressively deindustrializing (have China do the pollution for them, in fact do even more pollution, for the smuggies)
Italy, France and IIRC Spain (surprisingly) produce a small amount and every other country is a rounding error at best. Something that's a rounding error can still be economically important (look at low-impact legacy media, for example) but its not a dominant force.
You'll note almost all the "Green Steel" EU-funded sites are in non-steel producing areas, like Sweden. So it's one of those political football things where it's really not a "green" thing its more of a German empire (which is what the EU functionally in practice is) vs the outlying provinces where the non-German countries want to make some steel but central control wants to mercantile system the rest of the empire so there's lots of political back and forth about this, and very little if any engineering behind the decisions made.
Note its not even "green". Instead of using coal to make metallurgical coke to make steel, they turn enormous amounts of natgas (because .eu has so much natgas they don't have to import from Russia (sarcasm intended)) to make reformed hydrogen to make steel. They're going to emit the same amount or more or carbon, they're just not directly using coal onsite, so its magically "green". Carbon doesn't enter the building grounds of the steelmaking plant so its "green" even if they use more carbon than ever before as long as its used offsite. Its kind of lame as is the usual with greenwashing projects.
Despite enormous government handouts, they'll probably still find a way to make the green steel cost more than regular steel.
Its sort of the Euro's version of how the USA insists on adding ethanol to gasoline. Dumb, but politically driven and politically necessary.
(Score: 3, Interesting) by looorg on Thursday January 16, @04:35PM
Sweden produces steel, it's also the biggest Iron producer in the EU. So it would be odd if they didn't make steel. In that regard the EU produces more steel then the USA or India. It's just divided up by country. China makes about half, or slightly more, than the steel in the world. Second is EU as a whole, then India, Japan and the possibly the USA or Russia. So they all make steel in Europe, it's just minor amounts per country. It would be like saying which states in the USA makes the steel. Some probably make none, or close to. But a few make most of it.
(Score: 2) by VLM on Thursday January 16, @04:24PM
AI-generated "Tech Trends and Breakthroughs of 2026"
Oh wait I think "journalists" are already experimenting with this.
(Score: 2) by DannyB on Thursday January 16, @06:22PM
Challenge: try to use every last one of those buzz words in a single paragraph.
Stop asking "How stupid can you be?" Some people apparently take it as a challenge.