The latest version of Gemini is causing quite a stir in our engineering school. We've been testing it by snapping images of problems from exams on cell phones and asking Gemini to solve them. The worst exam performance we've seen so far is an overall grade of B-, and that includes graduate-level exams. Gemini even helpfully provides all of its intermediate work to show how it got the solution.

Gemini can even (correctly) find a solution for many simple design problems. We're scrambling to adapt to what's happening, and wondering what we'll be dealing with a year (or two) from now.

We had hoped that engineering (as opposed to computer science) would be relatively immune to AI disruption for at least a few more years, but we were very much mistaken. If all one evaluated was homeworks and exams, the best AIs are already capable of earning an engineering degree in most disciplines.

It turns out there's a big difference between you having data, and some US company having data. The government isn't tracking your phone. Private companies are. Hence LPR.

You're right, there's something magical about a cell phone provider that prevents an authoritarian government from simply demanding access to all location data at any time. Oh, wait ... there isn't.

My point being that if an authoritarian government does want to track everyone's location, LPR cameras would be only one tool in a much larger toolbox available to them - and frankly, not a particularly effective one.

The premise of this article makes no sense at all. Why would Uzbekistan need LPR cameras to track citizens when everyone is already carrying personal tracking devices in the form of smartphones? And unlike an LPR camera that only records when your vehicle is on the road, a smartphone reveals your personal location. Why even bother to query LPR camera records when the police can just ping the local cell phone provider and learn exactly where you are and where you have been for several months?

I see this same bizarre doublethink in my own city, where critics of LPR cameras decry their deployment while simultaneously recommending that everyone should be riding mass transit in the first place, where all passengers are placed under constant video and audio surveillance.

If you truly hate government surveillance, why do you tolerate it in mass transit? Read a little about the capabilities of the systems installed and maintained by March Networks in mass transit systems. LPR cameras are a joke compared to what our government leaders subject us to when we hop on a bus or a subway.

"I'm not sure a human driver would have avoided the dog either, though I do know that a human would have responded differently to a 'bump' followed by a car full of screaming people," the Waymo passenger wrote on Reddit.

I can tell you exsctly how many human drivers would respond in a situation like this, because I've seen it happen and have heard about it enough times: the driver would have accelerated away from the incident at high speed.

Lots of people are jerks. And others don't want to take the risk of confronting an angry (possibly armed) person who blames the driver for running over their pet.

The dog was unleashed. The legal fault lies with the owner. This was an unfortunate accident, but it is hardly Wayno's fault.

I'll counter your anecdote with my own. I know quite a few homeschooled kids who are now adults who are homeschooling their own children.

Religious motivations aside, those children can all read, write, and perform basic mathematics far above equivalent public school levels. Their grasp of basic science is pretty good, too, at least when it doesn't conflict with religious beliefs. And every one of them plays some sort of musical instrument, and is active in sports.

I was once a homeschooling skeptic myself, but the results speak for themselves. Done right, homeschooling provides excellent outcomes. Done poorly, not so much. But in that respect, it's certainly no different than a whole lot of public schools.

Shouldn't the ACLU and EFF be devoting their efforts to ending the Valley Transit Authority's comprehensive video surveillance of all its bus and light rail passengers first?

https://www.marchnetworks.com/...

https://www.youtube.com/watch?...

What's been going on with the VTA over the past twenty years makes the deployment of LPR cameras in San Jose look like a joke. An LPR system records license plates, but the VTA records comprehensive video and audio of every single passenger who rides on it. That includes where those passengers get on on off - exactly the sort of "tracking" that the ACLU and EFF condemn.

Exactly when did all those MTA passengers consent to give up their privacy? If privacy is sacrosanct for people driving on public roads, what about people on public transit?

"Woman Wrongfully Accused by a License Plate-Reading Camera" .... no, she wasn't.

LPR cameras don't accuse people, nor do they order their arrests. An LPR camera takes a picture of a license plate, reads it using standard image processing algorithms, and provides the plate number and plate image to law enforcement for verification.

What happens next depends on the police. This wasn't a failure of technology, but instead a failure of investigative procedure by the Columbine Valley Police Department.

It was also a failure of Ring cameras, which are absolutely abysmal for security purposes. As the article pointed out, the porch pirate was never recorded getting into the Rivian. That's because Ring cameras don't provide continuous-time internal recordings, since it would interfere with Ring's subscription model. Even a $50 Wyze Cam + memory card would have shown who got into and out of the vehicle.

The accused woman's Rivian was in the wrong place at the wrong time, and captured in the background during the recording of a crime. The CVPD added 2 plus 2 based on incomplete evidence, and got 5.

The correlation between gum / tooth disease and heart disease has been known for many years, but this study provides a mechanism for causation.

For that matter, talk to your veterinarian, and you'll be told that dogs with congestive heart failure and heart murmurs almost always have bad teeth.

I once agreed to participate in a medical study after undergoing a routine medical procedure. I was quite surprised at what I was asked, and how I was asked it.

It wasn't just a matter of self-reporting my information, but also that the survey designers presupposed that I would have a reliable memory of things I had done or eaten going back 6 months to several years, and could provide a meaningful average value. Some questions were asked repeatedly in slightly different ways, perhaps to see if my reponses deviated between them, but it did nothing but add to my annoyance.

About 30 minutes into what was supposed to be a 20-minute survey (with barely half the responses completed), I was shown multiple series of photographs where I had to choose the shade of cooked meat that I consumed, as an indicator of how well-done I liked it. The photos were so disgusting that I couldn't imagine touching anything in them. At that point I gave up.

The signal-to-noise ratio in this type of self-reported medical data must be incredibly low. I don't see how anyone could trust conclusions drawn from such studies.

These glasses and phone cameras are being *actively* used by a person staring at you in a place like a restaurant, which is just creepy, and there's a high likelihood that recording is going to be on Youtube or Tiktok, if they're not already streaming your dinner live.

You see the difference now?

I certainly see what you think the difference is, or perhaps what you think it should be. But give this technology five years, and you won't know you're being recorded, any more than you can know right now if the audio of a conversation you're having with a random stranger is being recorded.

Augmented Reality is coming, and soon you're going to see lots of people wearing AR-capable visual aids everywhere. You won't have a clue if the people using them are recording or not. So short of physically assaulting everyone you see in public venues, your choices are to avoid those venues, or learn to live with it.

Privacy is dead, courtesy of advanced technology that continues to spread at a breakneck pace. Pretending otherwise ignores the reality of how we all must adapt.

Video is easy to deal with. Just get some ultrabright infrared LEDs and wear them all over your body. The glassholes won't get anything but the glare from those.

Except that modern optical sensors with IR filters aren't affected in the least by this tactic.

The current trend in security cameras is "night time color". The cameras no longer go into IR-sensitive night time mode with IR illumination. They are sufficiently sensitive to operate even in very dim white light. All infrared sources are automatically filtered out. And when a particular technology becomes cheap enough to place in a sub-$100 security camera, it is just a matter of time before it speads everywhere.

The last thirty years of technological advancement have proven that if a technology is cheap, easy to use, provides some perceived advantage to the user, and can by rationalized by the user as causing no real harm to others, then it will be used .... everywhere.

Railing against smart glasses that record video makes as much sense, and is just about as effective, as railing against security cameras recording you, or your cell phone company always knowing your location, or Google and Amazon tracking everything you buy and every place you visit.

It is already trivial for someone to surrepticiously record the audio of any conversation on a smart phone or smart watch, then run it through a speech-to-text converter to produce a transcript. Smart glasses are just the next logical step.

This is the world we live in now. You must assume, always, that any interaction with another human being who you do not implictly trust is being recorded. We can't unmake the computers or the software, and when the same people who rail about their privacy then make a point of posting their concerns on social media to draw as much attention to themselves as possible, then they implicity undermine their own arguments.

I was never sure that they had the "best engineers", but they did have the best processes to make IC's.

They had (and still have) excellent engineers, but a terrible corporate reward structure. In my experience with them, engineers who worked on expensive high-margin products were rewarded and promoted. Everyone else was shoved aside.

Ever wonder why Intel made zero inroads into low-cost computing? They had the processes and the talent to do it, and even made an attempt with the (now cancelled) Galileo, Joule, and Edison product lines. But every time I contacted their Galileo group, a completely different group of engineers were running it. It was a dead-end career move at Intel, and every competent engineer tried to get out as quickly as possible.

Until and unless the internal corporate reward structure changes at Intel, the death spiral will continue.

Like most of the ills of academia, this one is largely self-inflicted.

For decades, promotion & tenure committees have held junior faculty to standards that they themselves could never have achieved. You'd have P&T committee members passing judgment on assistant professors who had published more journal papers than any three of them put together. It didn't matter - the bar was constantly raised.

So faculty increasingly turned to the MPU (minimum publishable unit) strategy - chopping up what should have been one really good paper into five or six mediocre ones. Even before AI exacerbated this problem, the crapflood of journal submissions was overwhelming reviewers and journal editors.

And now? It's become even more nightmarish. Many of my colleagues simply refuse to review papers any longer. They're done wasting their time going through ultra-dense text with perfect grammar and spelling that was clearly written by ChatGPT. There have even been Ph.D. students who attempted to pass their qualifying exams with immaculate presentations on material that they could not answer even the simplest questions about. So now we're moving into the second phase of the rot, where students who earned fraudulent Ph.D.'s become the next generation of faculty.

Academia will not be a pretty sight twenty years from now.

Tan said Intel will instead focus on "edge" AI that operates directly on devices rather than centralized computers.

And will those "edge" AI devices have to sell for $1000 each in order for the engineers who design them to be rewarded and promoted within the company?

That has been Intel's doom for the past three decades. The employees working on high-end products with high margin sales get all the attention and promotions. The ones working on high-volume, low margin products go nowhere. That's why Intel has made so little headway in anything beyond enterprise products in recent years.

This new "focus" will go nowhere unless Intel's corporate culture is vastly different than what is was when I last dealt with it.