That's hearsay bullshit, and not how justice works.
Let's be precise. "A says 'I saw Lance shooting up,'" is eyewitness testimony and is admissible. "A says 'B told me that he saw Lance shooting up,'" is hearsay and is not admissible. In a court of law, the prosecutor would probably decline to go with only the eyewitness testimony, unless it included enough to show that said injection was of banned drugs. It seems telling to me that several governments have conducted investigations and none have filed charges. Not that Lance is clean, but that there's insufficient evidence to file criminal charges. The USADA is a civil rather than criminal matter at best, and has a much lower evidenciary standard.
Let me qualify that. Simple, unobtrusive text ads, delivered at high speed so that they don't interfere with the rest of the user experience? Sure. What I've always noticed when I enable Adblock is how much faster the Internet feels.
Colorado is in the lowest sixth of US states for overall cancer rates.
This despite being in the top third for skin melanoma.
When you go in for a check-up,
the docs don't ask you whether you've checked the radon levels in your house.
But they will ask you if you wear sunblock,
and UV-blocking sunglasses (UV has been linked to cataract development).
Cause the UV levels that go with living at 5,000 feet are much more dangerous than the other radiation exposures.
The Comcast cable service my wife has us subscribe to for other content gave me access, so can't fuss about that. The NBC site prompted me for my cable provider the first time I tried to view a stream, and apparently got all the info they needed from my IP address. After the first access everything was transparent. The listing of when the live streaming of the (fencing) events I was interested in was accurate. Streams started promptly and played smoothly. Even for modest sized content (480p rather than high-def), decoding was compute-intensive, requiring the cycles from about 1.5 of the two processor cores on my Mac. That seemed excessive.
I'm a sport fencer. Epee if it matters. I wanted to watch the later rounds of the various epee events -- men's individual, women's individual, women's team. No men's team epee event at the Olympics this year, as the IOC has limited the number of fencing gold medals that can be won. None of the epee events were on US television, only available by streaming. Every minute of all the events were available, at least on replay. Except for one, the live events were either too early, or conflicted with the rest of my life. What was available in replay was the nearly raw video feed from the venue. The action, then a quick slow-motion replay of each touch. The director(s) obviously knew something about the sport, since the slow motion was generally the correct one of the two or three options for camera angle. Audio was the microphone for the referee of the bout being shown, plus ambient noise from the venue (including the PA). No announcer. No color analyst. No commercials. When the Koreans appealed the referee's decision and there was an hour of dead time from the venue, every minute of the dead time was included in the stream. As an aside for those who saw pictures of the Korean woman sitting on the strip, it wasn't a "protest" -- international fencing rules require the fencer to stay at the strip until the appeal is settled.
For an epeeist, that's really terrific coverage. I know what I'm looking for, and the announcer/color commentary are just a distraction. For a non-fencer, it must have been terrible.
If I were looking 20 years out (emphasis on at least that far in the future), my first questions would be whether the country/region has a high tech level broadly distributed today and is likely to have reliable local sources of water and electricity then. Yes, I know that puts me out on the lunatic fringe, but the arguments in Limits to Growth are still relevant and we're just now entering the period of interesting times in their forecasts. IMO, that rules out Africa, India, China, and most of the rest of Eastern Asia including Japan (in that order). I'd stay clear of isolated urban city-states (eg, Singapore). New Zealand seems a reasonable bet, also parts of Canada, Northern Europe, and select parts of the US. Brazil and Argentina if they can finish getting their acts together soon. Water may be in issue in Australia, but my main concern would be that they've got an awful lot of resources that China would like to more directly control. Canada may have that same problem with the US.
I gave my own kids the advice that within the US, they ought to stay in the contiguous states, roughly west of 105 W longitude, north of 40 N latitude.
From the perspective of someone putting on a sizable conference? World-class expertise at dealing with any unusual aspects of your conference. Daily inexpensive flights from almost everywhere. Plenty of inexpensive hotel space, and plenty of expensive space for the executives. By comparison, New York or San Francisco or Chicago are much more expensive. Venues guaranteed big enough to handle your crowd -- for things like CES and NAB, there's probably nowhere else in the country that can handle them. Off hours, entertainment ranging from free (a walk up and down the Strip at night is an experience -- tacky, perhaps, but unique) to expensive. Cheap cabs to and from everywhere, so there's no need to go through the hassles of a rental car. Gambling for the people who want it (I'm not a gambler, but I love watching strangers play blackjack).
If you know where to look, good food at reasonable prices. For example, don't know if it's still there, but there used to be a steakhouse buried in the depths of Circus Circus. Quiet, excellent service, very good food, very nice wine list, and didn't cost an arm and leg. But if you didn't know, you'd never find it by accident. And if you didn't make a reservation at least a couple of days in advance, you'd never get in if you did find it.
You'll also be aiming at a professorship, some research gig, or being woefully underworked and underpaid for your education until you've got enough hours in the trenches to make a good senior analyst or senior architect. [Emphasis mine]
My experience is that you mean overworked there.
Otherwise, spot on.
I might push some discrete math for code monkeys,
if only because it's a rich field for algorithms.
Liberal arts is not useless.... Math is a beautiful thing, it helps us explain the world.
So how much math are you going to require a liberal arts major to take?
There are academics writing serious pieces like this one in the New York Times
that suggest the proper amount of college-level math (calculus or above) to require is none.
Math as a language to describe the world would seem (at least to me) to require a minimum of
differential equations and a real statistics class,
both of which require a couple of semesters of calculus as a foundation.
Or you could simply require them to pass a couple of "topics" courses
to demonstrate that they understand there's a lot of different kinds of math out there --
but I certainly wouldn't trust them to actually use that math.
As to the question in the original post...
it depends on what kind of software you develop.
Much of mine was either real-time
(where algorithms for solving discrete math problems and the math for computational complexity were useful)
or tools to solve specific applied math problems
(where I needed to know the problem-specific math).
I always enjoyed his science fiction, but it always struck me as good, not great. OTOH, reading Three Hearts and Three Lions when I was a kid hooked me on fantasy forever.
At the very heart of the DOJ case...M$ was accused of "locking-in" customers for their products. And now, fast forward to 2012... Apple is literally locking in consumers behind their gardened walls with a plethora of their own hardware and software, Google & FB literally collecting private details from its consumers. Playing the devil's advocate here, I wonder how come they are not scrutinised intensely?
Such behavior is illegal only if you have a sufficiently large share of a properly defined product market. MS apparently got terrible legal advice in the 1990s (or ignored good advice); someone should have been telling them that they were dominant enough in their principle market space (personal computer operating systems) that the rules were different. Apple holds less than 20% of the global market for smartphones, a distant second behind Samsung for the most recently finished quarter. That's not enough market share to get you in trouble. Google appears ready to settle their antitrust case in the EU, and the FTC announced several months ago an investigation of Google's business practices in the US. And it's difficult to define an applicable "market" where Facebook dominates, since they don't charge their users.
Absolutely.
I've got boxes in the basement 40 years old with ball-point ink in cheap notebooks that are fine.
At that age some of the newsprint is getting rather yellowed,
but the ordinary paper and ink are okay.
Carbon black pigment typical of monochrome laser printers
on decent quality copy paper
is probably good for at least a century
if it's kept somewhere dark and dry.
My problems with digital media over the years has mostly been a matter of finding
equipment that can read the media and then decoding the application-specific bits.
If you have something that can handle the file system,
flat ASCII or Unicode text will almost certainly be recoverable;
open standards like PDF or Postscript are probably okay over 25 years;
a proprietary format may not be recoverable at all.
I've seen cases where recovering just the text from early Word documents is difficult,
and that's only been a little longer than 25 years.
Excel is more of a problem. For too much of the world, Excel is the default numerical computation platform because it can be assumed to be available. I'm not saying that Excel is a good platform, just that an enormous amount of the world uses it. And the Windows version has things that neither LibreOffice nor Office for Mac support consistently; eg, Solver and VBA. When Finance and the budget office say that their models and tracking tools require the Windows version of Excel, the decision about the company's standard spreadsheet and word processor has been made.
Some years back (circa 1990) the Phoenix airport shut down when the air temperature went above 120 degrees F, the maximum for which most jets had been certified (in terms of safe take-off weight) under US FAA rules. Several people from our company were stranded for a few hours until evening when the air temperature dropped back down. IIRC, Boeing took at least one model of all its jets to Saudi Arabia, along with the FAA-qualified measurement gear, and certified the planes to 125 degrees.
A few weeks earlier, the WSJ sparked a debate with its report that grammar gaffes have invaded the office in an age of informal e-mail....
You would think that the WSJ, being a leading business publication, would have made the point that once you get to court there's no such thing as informal e-mail. In that situation, it's not a hallway conversation, it's a discoverable document. And those messages can get you in serious trouble. For example, Microsoft's "cut off Netscape's air supply" enjoyed a prominent place in the judge's order (later overturned) to break up the company.
Can you turn your internal grammar checker on and off? Maybe it's because I'm getting to be an old fart, but I can't. And yeah, if I'm thinking in words, I think in grammatically correct sentences, or at least as close as I can come on the fly.
Too many people treat e-mail and text messages as if they were private conversations over the phone or in the hall. When someone sues you, though, the written words are discoverable documents. Statements that are ambiguous or easily misinterpreted can get you in real trouble. Microsoft almost paid dearly for the casual use of the phrase "cut off Netscape's air supply" in an e-mail. The trial judge used that quote to make a point in his order (later overturned) to break up the company.
From a grammar perspective, the last one, indicating a pause to suggest irony or a hidden meaning. Typographically, it depends on which style guide you use. The MLA's style guide, for example, insists on three dots with a full space separating the dots themselves and the text before and after. That's almost impossible to do in HTML, as most of the layout engines will strip out some of those spaces, even if you use the non-breaking space glyph. By the way, the plural is ellipses.
It's always a fun day when you arrive at work and the company lawyers have put yellow crime-scene tape across your doorway, along with a notice that they'll be spending the day going through your paper files and making a copy of your hard disk(s). There were two amusing incidents that day. They could copy the disk in the company-supplied Windows laptop without any problem; the headless Linux boxes in the little rack in the corner gave them fits. When they got to the locked file drawer where I kept stuff provided by potential vendors under NDA, I told them that they weren't on the list of people authorized to see those documents. The legal department spent all day resolving that one internally, and eventually decided to skip that drawer.
There are a couple of old stories told about the government's antitrust case against IBM (eventually dropped). In the first, the government subpoenaed a list of documents. IBM said they would comply, and a few weeks later three tractor-trailer rigs pulled up at the DOJ with copies of the documents, in boxes in random order. In the second, the DOJ decided that it needed to buy a document management system to organize all the documentation it received from IBM as part of the case. The only system available at that time that was big enough to handle the volume was sold by... IBM.
A bit faster at 15 fps would be better, but as you say, that's not a big jump.
When I was doing research on little video windows on computers 20 years ago, 13-15 fps was a critical speed because it was the rate at which people could determine if the audio and video were synced by watching lip movements. At 10-12 fps, the motion wasn't smooth enough to tell. For some applications, like some virtual classrooms where the video was a secondary medium -- the slides and the audio were the primary ones -- 15 fps video in a small window was adequate because what was important was the broad body language and overall facial expressions, not the fine detail.
Plastic Logic's website also describes work to allow them to incorporate a sensor layer along with the display. Heck with typing -- I'm willing to pay well for an electronic piece of paper that I can use to take math notes, sketch graphs, and so forth, at something approaching the size and resolution of paper and pen.
I've always emphasized to my calculus students that while calculus grew out of physics originally, from 1830-50 the mathematicians threw out that derivation and rebuilt it from the ground up... so it would be useful for things other than physics:^) Absent calculus, you can teach discrete probability; you can teach descriptive statistics for a sample; but you can't teach them continuous distributions and all that goes with that, other than cookbook approaches. OTOH, anyone who takes only one probability and statistics class is going to be dangerous, calculus or not: they're simply not going to be able to recognize all the things they don't know.
The results didnt seem too surprising, other than that under their questions, Visual Basic and Assembly ended up clustered together.
Odd things seem to happen when you change the number of clusters.
There seem to be too many cases where you start with N clusters and pick a pair of languages.
Then go to N+1 clusters and the pair is split between two clusters.
Then go to N+2 clusters and the pair is back in the same cluster.
While k-means is often the first tool one grabs for clustering analysis,
it isn't the only tool and
there are data sets where k-means doesn't give reasonable results.
The standard numbers that are tossed around for the average US suburban home (where a bit over 50% of the population lives these days), is 30 kWh per day, with a peak hour usage as high as 6 kWh, depending on location. IIRC, the peak hour tends to occur in the late afternoon and early evening, and varies somewhat between households: people coming home from school/work and turning on lights and A/C, parents firing up the washer/dryer, electric cooking, etc. We looked at converting to NG for cooking at one point; current code requires venting to the outside, which would in turn require some structural work. A contemporary single-family house in the US suburbs will be equipped with at least the equivalent of 125-amp 120-volt service (supports 15 kWh per hour max); the equivalent of 200-amp 120-volt service (24 kWh per hour) is not unusual.
Obligatory "Get off my lawn you damned kids!" anecdote. When my kids were in their early teens, I swear they could come through the front door and within 60 seconds, turn on 500 watts worth of assorted load each. Ever since, and after comparing notes with colleagues, I've claimed that one of the defining characteristic of dads who've had teenagers is a compulsive urge to turn things off and sit quietly in the dark.
Slashdot Mirror
Let's be precise. "A says 'I saw Lance shooting up,'" is eyewitness testimony and is admissible. "A says 'B told me that he saw Lance shooting up,'" is hearsay and is not admissible. In a court of law, the prosecutor would probably decline to go with only the eyewitness testimony, unless it included enough to show that said injection was of banned drugs. It seems telling to me that several governments have conducted investigations and none have filed charges. Not that Lance is clean, but that there's insufficient evidence to file criminal charges. The USADA is a civil rather than criminal matter at best, and has a much lower evidenciary standard.
Let me qualify that. Simple, unobtrusive text ads, delivered at high speed so that they don't interfere with the rest of the user experience? Sure. What I've always noticed when I enable Adblock is how much faster the Internet feels.
Here.
Colorado is in the lowest sixth of US states for overall cancer rates. This despite being in the top third for skin melanoma. When you go in for a check-up, the docs don't ask you whether you've checked the radon levels in your house. But they will ask you if you wear sunblock, and UV-blocking sunglasses (UV has been linked to cataract development). Cause the UV levels that go with living at 5,000 feet are much more dangerous than the other radiation exposures.
The Comcast cable service my wife has us subscribe to for other content gave me access, so can't fuss about that. The NBC site prompted me for my cable provider the first time I tried to view a stream, and apparently got all the info they needed from my IP address. After the first access everything was transparent. The listing of when the live streaming of the (fencing) events I was interested in was accurate. Streams started promptly and played smoothly. Even for modest sized content (480p rather than high-def), decoding was compute-intensive, requiring the cycles from about 1.5 of the two processor cores on my Mac. That seemed excessive.
I'm a sport fencer. Epee if it matters. I wanted to watch the later rounds of the various epee events -- men's individual, women's individual, women's team. No men's team epee event at the Olympics this year, as the IOC has limited the number of fencing gold medals that can be won. None of the epee events were on US television, only available by streaming. Every minute of all the events were available, at least on replay. Except for one, the live events were either too early, or conflicted with the rest of my life. What was available in replay was the nearly raw video feed from the venue. The action, then a quick slow-motion replay of each touch. The director(s) obviously knew something about the sport, since the slow motion was generally the correct one of the two or three options for camera angle. Audio was the microphone for the referee of the bout being shown, plus ambient noise from the venue (including the PA). No announcer. No color analyst. No commercials. When the Koreans appealed the referee's decision and there was an hour of dead time from the venue, every minute of the dead time was included in the stream. As an aside for those who saw pictures of the Korean woman sitting on the strip, it wasn't a "protest" -- international fencing rules require the fencer to stay at the strip until the appeal is settled.
For an epeeist, that's really terrific coverage. I know what I'm looking for, and the announcer/color commentary are just a distraction. For a non-fencer, it must have been terrible.
If I were looking 20 years out (emphasis on at least that far in the future), my first questions would be whether the country/region has a high tech level broadly distributed today and is likely to have reliable local sources of water and electricity then. Yes, I know that puts me out on the lunatic fringe, but the arguments in Limits to Growth are still relevant and we're just now entering the period of interesting times in their forecasts. IMO, that rules out Africa, India, China, and most of the rest of Eastern Asia including Japan (in that order). I'd stay clear of isolated urban city-states (eg, Singapore). New Zealand seems a reasonable bet, also parts of Canada, Northern Europe, and select parts of the US. Brazil and Argentina if they can finish getting their acts together soon. Water may be in issue in Australia, but my main concern would be that they've got an awful lot of resources that China would like to more directly control. Canada may have that same problem with the US.
I gave my own kids the advice that within the US, they ought to stay in the contiguous states, roughly west of 105 W longitude, north of 40 N latitude.
From the perspective of someone putting on a sizable conference? World-class expertise at dealing with any unusual aspects of your conference. Daily inexpensive flights from almost everywhere. Plenty of inexpensive hotel space, and plenty of expensive space for the executives. By comparison, New York or San Francisco or Chicago are much more expensive. Venues guaranteed big enough to handle your crowd -- for things like CES and NAB, there's probably nowhere else in the country that can handle them. Off hours, entertainment ranging from free (a walk up and down the Strip at night is an experience -- tacky, perhaps, but unique) to expensive. Cheap cabs to and from everywhere, so there's no need to go through the hassles of a rental car. Gambling for the people who want it (I'm not a gambler, but I love watching strangers play blackjack).
If you know where to look, good food at reasonable prices. For example, don't know if it's still there, but there used to be a steakhouse buried in the depths of Circus Circus. Quiet, excellent service, very good food, very nice wine list, and didn't cost an arm and leg. But if you didn't know, you'd never find it by accident. And if you didn't make a reservation at least a couple of days in advance, you'd never get in if you did find it.
My experience is that you mean overworked there. Otherwise, spot on. I might push some discrete math for code monkeys, if only because it's a rich field for algorithms.
So how much math are you going to require a liberal arts major to take? There are academics writing serious pieces like this one in the New York Times that suggest the proper amount of college-level math (calculus or above) to require is none. Math as a language to describe the world would seem (at least to me) to require a minimum of differential equations and a real statistics class, both of which require a couple of semesters of calculus as a foundation. Or you could simply require them to pass a couple of "topics" courses to demonstrate that they understand there's a lot of different kinds of math out there -- but I certainly wouldn't trust them to actually use that math.
As to the question in the original post... it depends on what kind of software you develop. Much of mine was either real-time (where algorithms for solving discrete math problems and the math for computational complexity were useful) or tools to solve specific applied math problems (where I needed to know the problem-specific math).
I always enjoyed his science fiction, but it always struck me as good, not great. OTOH, reading Three Hearts and Three Lions when I was a kid hooked me on fantasy forever.
Such behavior is illegal only if you have a sufficiently large share of a properly defined product market. MS apparently got terrible legal advice in the 1990s (or ignored good advice); someone should have been telling them that they were dominant enough in their principle market space (personal computer operating systems) that the rules were different. Apple holds less than 20% of the global market for smartphones, a distant second behind Samsung for the most recently finished quarter. That's not enough market share to get you in trouble. Google appears ready to settle their antitrust case in the EU, and the FTC announced several months ago an investigation of Google's business practices in the US. And it's difficult to define an applicable "market" where Facebook dominates, since they don't charge their users.
Absolutely. I've got boxes in the basement 40 years old with ball-point ink in cheap notebooks that are fine. At that age some of the newsprint is getting rather yellowed, but the ordinary paper and ink are okay. Carbon black pigment typical of monochrome laser printers on decent quality copy paper is probably good for at least a century if it's kept somewhere dark and dry.
My problems with digital media over the years has mostly been a matter of finding equipment that can read the media and then decoding the application-specific bits. If you have something that can handle the file system, flat ASCII or Unicode text will almost certainly be recoverable; open standards like PDF or Postscript are probably okay over 25 years; a proprietary format may not be recoverable at all. I've seen cases where recovering just the text from early Word documents is difficult, and that's only been a little longer than 25 years.
Will Office finally admit that basic 100-year-old typesetting functions such as floating displays should be supported?
Excel is more of a problem. For too much of the world, Excel is the default numerical computation platform because it can be assumed to be available. I'm not saying that Excel is a good platform, just that an enormous amount of the world uses it. And the Windows version has things that neither LibreOffice nor Office for Mac support consistently; eg, Solver and VBA. When Finance and the budget office say that their models and tracking tools require the Windows version of Excel, the decision about the company's standard spreadsheet and word processor has been made.
Some years back (circa 1990) the Phoenix airport shut down when the air temperature went above 120 degrees F, the maximum for which most jets had been certified (in terms of safe take-off weight) under US FAA rules. Several people from our company were stranded for a few hours until evening when the air temperature dropped back down. IIRC, Boeing took at least one model of all its jets to Saudi Arabia, along with the FAA-qualified measurement gear, and certified the planes to 125 degrees.
You would think that the WSJ, being a leading business publication, would have made the point that once you get to court there's no such thing as informal e-mail. In that situation, it's not a hallway conversation, it's a discoverable document. And those messages can get you in serious trouble. For example, Microsoft's "cut off Netscape's air supply" enjoyed a prominent place in the judge's order (later overturned) to break up the company.
Can you turn your internal grammar checker on and off? Maybe it's because I'm getting to be an old fart, but I can't. And yeah, if I'm thinking in words, I think in grammatically correct sentences, or at least as close as I can come on the fly.
Too many people treat e-mail and text messages as if they were private conversations over the phone or in the hall. When someone sues you, though, the written words are discoverable documents. Statements that are ambiguous or easily misinterpreted can get you in real trouble. Microsoft almost paid dearly for the casual use of the phrase "cut off Netscape's air supply" in an e-mail. The trial judge used that quote to make a point in his order (later overturned) to break up the company.
From a grammar perspective, the last one, indicating a pause to suggest irony or a hidden meaning. Typographically, it depends on which style guide you use. The MLA's style guide, for example, insists on three dots with a full space separating the dots themselves and the text before and after. That's almost impossible to do in HTML, as most of the layout engines will strip out some of those spaces, even if you use the non-breaking space glyph. By the way, the plural is ellipses.
It's always a fun day when you arrive at work and the company lawyers have put yellow crime-scene tape across your doorway, along with a notice that they'll be spending the day going through your paper files and making a copy of your hard disk(s). There were two amusing incidents that day. They could copy the disk in the company-supplied Windows laptop without any problem; the headless Linux boxes in the little rack in the corner gave them fits. When they got to the locked file drawer where I kept stuff provided by potential vendors under NDA, I told them that they weren't on the list of people authorized to see those documents. The legal department spent all day resolving that one internally, and eventually decided to skip that drawer.
There are a couple of old stories told about the government's antitrust case against IBM (eventually dropped). In the first, the government subpoenaed a list of documents. IBM said they would comply, and a few weeks later three tractor-trailer rigs pulled up at the DOJ with copies of the documents, in boxes in random order. In the second, the DOJ decided that it needed to buy a document management system to organize all the documentation it received from IBM as part of the case. The only system available at that time that was big enough to handle the volume was sold by... IBM.
A bit faster at 15 fps would be better, but as you say, that's not a big jump.
When I was doing research on little video windows on computers 20 years ago, 13-15 fps was a critical speed because it was the rate at which people could determine if the audio and video were synced by watching lip movements. At 10-12 fps, the motion wasn't smooth enough to tell. For some applications, like some virtual classrooms where the video was a secondary medium -- the slides and the audio were the primary ones -- 15 fps video in a small window was adequate because what was important was the broad body language and overall facial expressions, not the fine detail.
Plastic Logic's website also describes work to allow them to incorporate a sensor layer along with the display. Heck with typing -- I'm willing to pay well for an electronic piece of paper that I can use to take math notes, sketch graphs, and so forth, at something approaching the size and resolution of paper and pen.
I've always emphasized to my calculus students that while calculus grew out of physics originally, from 1830-50 the mathematicians threw out that derivation and rebuilt it from the ground up... so it would be useful for things other than physics :^) Absent calculus, you can teach discrete probability; you can teach descriptive statistics for a sample; but you can't teach them continuous distributions and all that goes with that, other than cookbook approaches. OTOH, anyone who takes only one probability and statistics class is going to be dangerous, calculus or not: they're simply not going to be able to recognize all the things they don't know.
Odd things seem to happen when you change the number of clusters. There seem to be too many cases where you start with N clusters and pick a pair of languages. Then go to N+1 clusters and the pair is split between two clusters. Then go to N+2 clusters and the pair is back in the same cluster. While k-means is often the first tool one grabs for clustering analysis, it isn't the only tool and there are data sets where k-means doesn't give reasonable results.
The standard numbers that are tossed around for the average US suburban home (where a bit over 50% of the population lives these days), is 30 kWh per day, with a peak hour usage as high as 6 kWh, depending on location. IIRC, the peak hour tends to occur in the late afternoon and early evening, and varies somewhat between households: people coming home from school/work and turning on lights and A/C, parents firing up the washer/dryer, electric cooking, etc. We looked at converting to NG for cooking at one point; current code requires venting to the outside, which would in turn require some structural work. A contemporary single-family house in the US suburbs will be equipped with at least the equivalent of 125-amp 120-volt service (supports 15 kWh per hour max); the equivalent of 200-amp 120-volt service (24 kWh per hour) is not unusual.
Obligatory "Get off my lawn you damned kids!" anecdote. When my kids were in their early teens, I swear they could come through the front door and within 60 seconds, turn on 500 watts worth of assorted load each. Ever since, and after comparing notes with colleagues, I've claimed that one of the defining characteristic of dads who've had teenagers is a compulsive urge to turn things off and sit quietly in the dark.