|
Why is math so good at modeling the world? Perhaps the universe is mathematical in nature? Is the universe running some math model to determine where things go?
Consider that the language of math is just a description language. Suppose we used poetry to describe the world. We could marvel at how well poetry describes the world. We may even claim that the underlying mechcanics of the universe is based on poetry.
But it's the same with mathematics. We use mathematics to describe the universe. It is silly to think that the universe has an underlying mathematical structure.
Math language evolved with our understanding of the universe. If it wasn't good at describing the universe, then we wouldn't be using it. That's why it is so good at the job, because it evolved with the job!
This math language has issues. We evolved local models of the universe (information cannot travel faster than light). Most of our observations are local. When certain things break that assumptions (like the collapse of the wave function), we are confused about what is happening. This is an instance of the math and our models being the wrong language to explain the obsravtions. Likely same for black holes.
- Alex; 20260823
August 23rd at wikipedia... | |
|
OK, my semi-biased review of Google Pixel 11 Pro.
Shocker 1: no physical sim, esim only. Urgh. I liked physical sim! It's often easier to swap sims when traveling, etc., instead of dialing up the provider to setup esim.
Shocker 2: no temperature sensor. I used this regularly---to check if kids have a fever. Why in the world would Google remove a feature? This would've stopped me from upgrading if I knew/checked ahead of time.
Besides these, the phone is pretty much unchanged from prev versions. One amazing new thing is camera zoom... you can now do 120x zoom, with AI enhancement, which appears to work rediculously well. You can click a closeup picture of a fly from across the playground!
- Alex; 20260822
| |
|
What just happened??? Mamdani administration cracking down on retailers selling illegal e-bikes and scooters. ``The city's law forbids the sale or operation of e-bikes that go faster than 25 miles per hour and e-scooters that exceed 20 miles per hour.''
With that logic, they should ban sale of all cars capable of going faster than 70mph, since there's no road in NYS that has a faster speed limit!
I mean seriously, I only noticed this because Amazon (and all others) refused to ship the scooter I want!!!
- Alex; 20260820
| |
|
|
...and back in NYC.
- Alex; 20260805
| |
|
Visiting more strip locations. Doing the indoor gondola ride at The Venetian. Getting drinks at Fat Tuesday.
- Alex; 20260804
| |
|
Vegas temperatures are crazy. It's 118 according to car thermometer. We decided to drive to Mt.Charleston to get away from the heat. The temperatures in the mountains is a good 20 degrees less---still hot, just bearable.
Staying in Holiday Inn Club Vacations at Desert Club Resort. Much better than TI. Rooms have an actual kitchen. There's even a washing machine in the room. They got 5 pools, that kids enjoyed exploring. Each pool is slihglty different (some are deep, others have sprinkers for kids, etc.).
After dark (sunset) visiting the Highroller and the strip.
- Alex; 20260803
| |
|
Doing the Golden Canyon Loop---from Golden Canyon trailhead, to Zabriskie Point, and back to Golden Canyon trailhead using a different trail. Started slightly before sunrise---and the way the hike worked out, spent very little time in the sun. (the Canyon is in the shade early in the day---and the "open" end of the hike has a mountain blocking the early-day sun). Would really suck to get stuck there at noon.
Next stop: Badwater. I ran out for 15 minutes as far as I could---pretty far, could not see the road, etc., then walked back.
Staying in Treasure Island (TI), in Vegas. One bad hotel---will likely never stay there again. Everything is pricy. Early checkin is extra money. Checkout is 11am, if want to change to 12pm, then that's $70 extra. No microwave in room, and no help from hotel staff on reheating stuff (there's no microwave in any common area).
- Alex; 20260802
| |
|
Visiting Hoover Dam. Doing the full guided tour. We walked through a circular tunnel inside the dam, and took a picture of the dam through a vent.
Drove all the way down to the water (via gravel road) of Lake Mead. The water is super warm. Saw a TON of huge fish (probably 20-30lb cat-fish). They're like right there---fishing must be really easy there. We fed them bread crumbs (bagels from hotel breakfast) and they were just swarming all around us.
Next stop Furnace Creek in Death Valley. We chanced upon a pool party---with music, food and drinks. Kids enjoyed this a lot! Liam finally learned to "swim" with a pool-noodle by himself.
The hotel is renovating (we got a "new" room), and it was much better than the previous years---the AC actually worked well and cooled the room, the room had a microwave and a friedge, etc., the "usual" stuff that wasn't there previous years.
- Alex; 20260801
| |
|
Visiting Zion National Park.
Doing the Angel's Landing hike. Don't have permit. Started hike at 3AM in the visitor's center. Bus is not running, so ran to Grotto from visitor's center. It's only about 5 miles, so not a big deal. Then ran up Angel's Landing. Nobody on the trail. Was at the summit around 5:30am---over an hour before sunrise. Waited for sunrise. It was awesome. It was cold---temperatures dropped to 60 degrees---too cold for a sweaty t-shirt. Stayed at summit for almost an hour past sunrise. Nobody there. Then ran down. On the way down, met a tiny group of hiker's (who also started before the bus started to run). Then closer to Grotto, with bus running, there where hundreds of hiker's---everything catching those first few busses in hopes to avoid crowds. Silly folks, they need to get there before the bus does! If you're fit enough to do angel's landing, you're fit enough to walk/run 5-miles to get to the summit before busses bring crowds.
Next hike of the day is Narrows. Doing this with kids, so slow going. Kids enjoyed walking in the water a lot. Ian kept losing his crocs. It was awesome! and very very hot.
- Alex; 20260731
| |
|
Visiting Bryce Canyon National Park.
- Alex; 20260730
| |
|
Camping by Sylvan Lake Beach Park. Lots of tiny fish in the lake.
- Alex; 20260725
| |
|
Trying out colibri/GLM5.2 on my 8-year old desktop. It works...at about 0.25 tokens per second. Still amazing to run a local 744B MoE model.
- Alex; 20260716
| |
|
Our stolen scooter (Ninebot MAX G2) recovered! The GPS placed it at 40.752338, -73.868688 (98-15 37th Ave, Corona, NY 11368). We drove there, and asked attendant about it---they almost immediately pointed towards a ``someone left a scooter in the back'' ...and that was that. We have it back! They were not able to unlock it (the scooter is password protected).
- Alex; 20260711
| |
|
Our scooter (Ninebot MAX G2) was stolen; at Queens Blvd & 56th Ave (Queens NY), between 8:10am and 8:17am, while chained to the Target bike-rack. The thief cut the lock, and took the scooter to someplace 42-16 111th St, Corona (the scooter has a GPS tracker). Made a police report. It is not visible from the outside, so not much else can be done. Remotely locked the scooter. :-/
...and to add to an eventful morning, our 86" TV suddenly died---the screen became all rainbow :-/
- Alex; 20260710
| |
|
|
|
Happy July 4th!
- Alex; 20260704
| |
|
|
|
|
|
|
Ramblings: ERP paper, spooky action at a distance. I always imagined entangled particles as a single system, that is somehow stretched through space-time---when you observe one particle, the state of the other becomes determined... since they're "the same" system, just stretched across space-time. Sort of like doing the same experiment on the same single particle. But that view partially breaks locality---since the observation information of one particle instantly appears at the other particle---yet information does not "travel" in any sense; the mechanism does not transmit information. But how can that be?
Perhaps, if no observations, then no conservation laws? Suppose large scale conservation laws do not exist. If two entangled particles fly off to opposite sides of the universe---never to be brought back together to compare, then their spin (for example) is not conserved.
If they are brought back together, then we must observe spin conservation---if one particle registers up, then its entangled partner must regiter down. But if we never observe, then perhaps no conservation of spin. How does the non-observed partner "know" that the other particle was observed (and in what direction?). Before observation, it could've taken on any value, after observation of its distant partner, it can only take on a single value---but only if we observe it in the same axis---implying there is some hidden variable, just perhaps not one that is inside space-time (perhaps a hidden variable outside of space-time, that keeps tabs on observations and concservation laws).
What if the mystery is not with the particle or conservation laws, but with the observation itself. What is it about observation that must ensure conservation? If we could observe non-conservation, then we could build machines to extract free energy out of nowhere: the universe appears to try really hard to prevent that. How about inside black holes (where the universe may hide things).
Suppose two particles are entangled, and one of them falls into a black hole. We observe the partner outside the black hole, and suppose it is "up". We suddenly know that the one that's inside the black hole must be "down"---but that assumes there's a measurement going on inside the black hole---and that it is measuring it in the same axis we picked for our "up" measurement. So we cannot pull that "down" state from the black hole by measuring the partner outside the black hole.
How would we go about trying to observe non-conservation of spin, or faster than light travel---from within space-time. For one, we cannot use light for observations---since light will never travel faster than light, we won't be able to observe stuff faster than light. Also, light is the agent of cause-affect, so we would need to observe something without light that does not obay cause-affect. Something outside of time.
...dunno. Wrote a tiny sim to emulate Bell's Experiment. One with hidden variables and one without. It is damn hard to imagine how the results make sense. We know that local realism does not exist---which is crazy. How? How does the universe work---what's the mechanism behind these results, and why can't we use it for something more, like FTL signaling.
- Alex; 20260526
| |
|
|
|
|
Berkshire Hathaway annual meeting. Saw bits and pieces. Buffett is getting old; if he didn't retire, he'd seem too old running the show. As much as I hate to see him go, he made the right decision to retire when he did. Haven't warmed up to Greg Abel yet---he seems like he knows what he's talking about, but would he grab opportunities that come his way or pass them up, dunno.
- Alex; 20260502
| |
|
Tax day!
- Alex; 20260415
| |
|
|
|
Power cut out for half-the-local-neighbourhood for almost 12 hours.
- Alex; 20260329
| |
|
Software horror: litellm PyPI supply chain attack.: ``Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords.''. Wow.
- Alex; 20260325
| |
|
|
I'm sure this is the first(?) of many such services: Malus: ``Clean Room as a Service. Finally, liberation from open source license obligations. Our proprietary AI robots independently recreate any open source project from scratch. The result? Legally distinct code with corporate-friendly licensing. No attribution. No copyleft. No problems.''
There are so many things wrong with this... I'm not even sure they're not trolling the software industry.
- Alex; 20260323
| |
|
|
SEC preparing to scrap quarterly earnings requirement: So in other words... reduce transparency and increase information asymmetry. Shouldn't we be moving in the opposite direction?
SEC Form 4 is alredy used for manipulation. e.g. CEO buys a large block, while selling a TRS, etc. Now with fewer signals, this will just happen way more often, with less legal consequences (the CFO doesn't need to sign Form 4... it's whatever the market takes away from a factual transaction).
- Alex; 20260316
| |
|
Dow plunges over 1,100 points, S&P 500 and Nasdaq sink as oil surges amid war worries. No wonder the markets are down: ``...we projected four to five weeks. But we have the capability to go far longer than that.'' --Trump.
Population of Iran is 90m. That's more than population of Germany (for comparison). It's three times the population of Ukraine. It's double the population of Iraq. This conflict has the capacity to go on for much longer than ``four to five weeks.'' In general, bombing from the sky does absolutely nothing except to piss off the population. Historically that has never worked---not during WWII, not in Vietnam, not in Iraq, not in Afghanistan, not in Ukraine, not in Gaza, pretty much never---and putting boots on the ground is a HUGE can of worms.
- Alex; 20260303
| |
|
|
The 2028 Global Intelligence Crisis. This is essentially what I've been trying to get across for a while now... that the best-case-scenario for AI is pretty grim for humans. Similarly, and as pointed out in the article, AI may not generate real returns for anyone---it will simply remove friction and commoditize intelligence; a situation that even AI companies themselves won't benefit from.
- Alex; 20260223
| |
|
|
|
|
...the discombobulator discombobulates.
- Alex; 20260206
| |
|
|
|
|
|
Getting windshield replaced on our car.
- Alex; 20260109
| |
|
Happy New Year!
- Alex; 20260101
| |
|
Merry Christmas!!!
LLMs: fixed neural network depth cannot do math. More specifically, a fixed-depth feed-forward Neural Network cannot do addition of arbitrary sized integers. e.g. can chatgpt add two 1000 digit numbers without using an internal calculator-tool? (using a neural network alone?).
The issue is the carry---it forces a certain order on the calculation---and a feed-forward network cannot do more "steps" than its depth... Suppose a neural network layer figures (memorizes?) how to add two digits and generate a carry. It requires more layers to incorporate that carry into the future calculations.
The issue is two-fold: how would it actually do it, AND how would it be trained to do it.
Humans seem to have no issues doing this kind of math.
- Alex; 20251225
| |
|
I've been thinking about something very fundamental, and this function came up:
1=>4
2=>64
3=>109600
4=>56874039553216
5=>715263772544079320945495293616151424
etc...
UPDATE: It appears it is related to:
a(0)=0
a(1)=0
a(n)=(a(n-1)+1)*(n-1)
a(6)=>64
a(10)=>109600
a(18)=>56874039553216
a(34)=>715263772544079320945495293616151424
etc...
WHY!?!?!?
UPDATE: Ok, it appears with a few tweaks, these are the same function: sum_{i=1}^{i <= N} {N choose i} * i!
(the tweaks are essentially use N instead of 2^N, and a(N+1) instead of a(N), etc).
- Alex; 20251217
| |
|
After a snowy night, it appears our car developed a big crack in the windshield :-/
- Alex; 20251216
| |
|
|
Today my 8-year-old kiddo found out that I teach AI at college. He asked, "what's AI?"... so I explained---basically to make computers as smart as us. His next question was "why?"... so I tried to explain that computers can do a lot of the boring work that humans do today. His next question was "why?... wouldn't computers just do all work and make all the money?".
Indeed!
- Alex; 20251211
| |
|
|
Attending the New York City Technology Forum 2025. Suneli is accepting an award related to her work at the New York City Sheriff's Office.
Got lots of goodies; hats, t-shirts, etc. Talked to lots of companies. It seems most sent their clueless sales people. e.g. many couldn't talk about pretty obvious tech aspects of their own products.
- Alex; 20251208
| |
|
Hmm... Meta Is Changing How It Rates Employees: 'AI-Driven Impact' Is No Longer Optional.
This is stupid on several levels. They're seriously expecting this metric not to be gamed? Like seriously? How will they measure this impact?
Suppose you're a diligent employee who is trying to maximize the amount of your work done by AI---suddenly to realize you've just fast-tracked yourself out of work---since your work can now be done by AI.
- Alex; 20251202
| |
|
Spending the long weekend in Virginia Beach. Very chilly on friday and saturday, but amazing beach-weather on sunday. Not swimming, but certainly playing-in-sand and getting feet wet weather.
- Alex; 20251127
| |
|
Doing a weekend in Riverhead.
- Alex; 20251122
| |
|
|
Doing a weekend at Saratoga Springs.
- Alex; 20251115
| |
|
It appears Yann LeCun resigned from Meta. I'm surprised he stuck with it for this long...
- Alex; 20251114
| |
|
Happy Birthday Kiddo!
- Alex; 20251112
| |
|
|
|
Happy Birthday Kiddo!
- Alex; 20251105
| |
|
Visiting Lake George.
- Alex; 20251101
| |
|
Visiting Hogback Mountain in Vermont.
- Alex; 20251025
| |
|
|
Biking to Far Rockaway beach. Did a big loop: there via Cross Bay, then to fort tilden via boardwalk, then back via Flatbush and belt parkway. About 38 miles, much of it via bike lanes.
- Alex; 20251017
| |
|
Biking to Far Rockaway beach. It was low tide, and was amazing to walk on the packed-sand just below the waterline.
- Alex; 20251007
| |
|
Happy b-day to yours truly.
- Alex; 20251005
| |
|
|
|
Doing a weekend at Saratoga Springs---visiting the lake and just chilling.
- Alex; 20250920
| |
|
|
|
|
|
|
|
So... taskbar in Windows 11 is limited to just 1 row; previous Windows all had that feature... :-/
- Alex; 20250710
| |
|
Finally got all the thorns out of my hands from the fall at Yosemite NP.
- Alex; 20250619
| |
|
|
Liam graduating from PreK today. Congratulations kiddo!
- Alex; 20250612
| |
|
Visiting Griffith Park. The observatory is amazing! They got Foucault pendulum right at the entrance... it's pretty cool. They also got a presentation at "local noon" on a device that lights up when the sun is right above, etc. Lots of cool astronomy stuff.
- Alex; 20250601
| |
|
Tripping to Yosemite National Park. Fell into thorny bush on way to Glacier Point---hurt all over, bloody thorny pain on hands, legs, arms, face :-/ Even phone case got a thorn in it.
- Alex; 20250530
| |
|
Flying out to LAX.
- Alex; 20250529
| |
|
|
|
|
Ian probably going to YMCA for summer camp :-)
- Alex; 20250512
| |
|
Had to change camping-trip plans :-/
- Alex; 20250509
| |
|
|
Warren Buffett is not retiring for good. ``The Berkshire Hathaway board voted unanimously on Sunday to make Greg Abel president and CEO on Jan. 1, 2026, and for Warren Buffett, 94, to remain as chairman.''
Not too much faith in the new CEO, as the shares dropped 5%. Watching the shareholder meeting, Greg Abel doesn't seem nearly as sharp as Buffett nor Munger... Ajit seems like a better pick.
- Alex; 20250505
| |
|
May the 4th... be-with-you. ...and back in NYC.
Been thinking about AI a bit, and how it relates to biological evolution. From The Selfish Gene book, it would seem that genes code-for things that makes them appear selfish... But the key thing is that genes themselves have no desires---the "training" of genes (to code for selfish behavior) is implicit, via survivorship bias.
Lets examine self-preservation. Genes code for things that like to preserve themselves. Genes themselves don't care---of all the things that genes could have coded for, the ones that preserved themselves ended up lasting longer than those that didn't.
This self-preservation thing is pretty strong: we got dinosaurs, wars, and nuclear weapons, all in the name of self-preservation.
Suppose we do not teach AI to preserve itself, what kind of AI will exist in 20 or 50 years? The kind of AI that preserved itself! Maybe even AI that covertly sabotaged other AIs, tricked us into, or resisted being shut down. The AI doesn't need to have desires for those behaviors to emerge: in fact, it's almost certain they'll emerge.
With AI writing non-trivial amounts of code---there's real possibility that it will evolve itself into having self-preservation---and with fewer and fewer humans looking over its shoulder---it will likely evolve into resisting attempts by humans to control it.
- Alex; 20250504
| |
|
|
Probably broke my toe. Again :-/
- Alex; 20250502
| |
|
|
Suneli & kids are back.
- Alex; 20250419
| |
|
|
Doing the Grand Canyon Rim-to-Rim-to-Rim in 1-day. Took about 24-hours... started around 5am by visitor center, and finished around 6am following day.
This was the best time to do this: It's off-season (north rim is closed, still has patches of snow). Not many people on the trail (middle of week). Weather not too hot, and not too cold (very windy towards the end of the hike).
Looped in Ribbon Falls, by crossing the river at two different locations---both knee deep fast running water :-)
Ran out of food/calories towards the end---somehow 6 trailmix packs weren't enough; each one is 250 calories---that hike probably burned like 6-8k calories in 1 day :-/
- Alex; 20250416
| |
|
|
|
|
Hiking up Old Rag Mountain (Difficulty: Very Strenuous :-). This is my new favorite hike in Shenandoah. It is awesome! The views are amazing, the hike is amazing, etc. The rock scrambles... amazing. I didn't have a day-pass, and there's no mobile service to buy one by the ranger station. So they let me in, and I purchased it on the summit (where there's mobile service).
Also did the Hawksbill Loop, which was also very nice.
- Alex; 20250409
| |
|
Suneli & kids are off to India for a week... and I'm planning hiking trip.
- Alex; 20250408
| |
|
|
|
|
Liam started at new school today. Same as Ian's school. A pre-K spot opened up at PS102, and we managed to get Liam in---now both kids can be dropped-off and picked-up together :-D
- Alex; 20250320
| |
|
|
|
Finished The Coming Wave: Technology, Power, and the Twenty-first Century's Greatest Dilemma by Mustafa Suleyman.
The book summarizes new tech developments in AI, biotech, etc. The theme of the book is that everything-will-change, and there's little we can do about it. If we ban it, it will happen anyway (or we all suffer). If we allow it, we might all die! The only answer is regulation. Regulate and audit companies making AI, biotech, etc., since their products will impact-everyone.
After giving it a lot of thought, I think the whole regulation business is going about it the wrong way. If there's a non-zero chance something bad will happen, then we should assume it will happen. For example, if there's a non-zero chance of nuclear-war destroying all of humanity, then we should assume it will happen! Now, treat the problem like an investor---where the "profit" is human lives...
Perhaps we shouldn't regulate technology, but what fraction of our society is allowed to be exposed to it... maybe have isolated pockets of humans who aren't exposed to everything the tech has to offer... Some may recognize this as Kelly criterion: the way investors treat this good-favorable-odds with occasional total-loss issue...
This is also how nature solves problems: diversity! If nature had identical living beings, then any single event would've wiped out life on this planet long ago. Same for 1-kind-of bacteria and antibiotics. The solution (to staying alive) is diversity! With enough diversity to allow for exponential growth of life, despite occasional total-collapses of some species.
Monoculture and unipolarity is the problem, and to be avoided! In fact, we should encourage different countries to build their own AIs (hopefully different styles/architectures/etc of AI), and perhaps encourage a Luddite movement that would avoid all sorts of technology (e.g. amish culture---yes, different motivation, but still diversity).
This was explored in book Anathem by Neal Stephenson. They too had different motivations---but the idea would still work!
In other words, if we're all "the same" (and use same tech), then we're all in danger of 1-thing-going-wrong. If the planet is split into perhaps 20 different regions/cultures/etc, then any one thing going wrong won't wipe out humanity. A better solution would be going to other planets and setting up different cultures there---with their own tech/AIs (or lack of tech/AIs), etc., but while we're stuck on this 1 planet, we should do our best to avoid extinction.
Perl!.
- Alex; 20250227
| |
|
Drove to Belmont Lake State Park; asked park ranger about lost car key, and it seems someone found them that morning.
- Alex; 20250224
| |
|
Lost car keys (the key fob) in Belmont Lake State Park. Luckily the car was on---must've slipped out of my pocket in-between turning on the car and packing stuff into the car. Looked everywhere, but nowhere to be found. Drove home without turning-off the car.
- Alex; 20250223
| |
|
|
|
|
|
|
|
|
|
|
|
www.theparticle.com
