11 ms·

> And I see the same trend in Machine learning, Big data, AI, Cryptocurrency, Internet of Things, Social networks, Decentralized networks, etc. Everything in li

> And I see the same trend in Machine learning, Big data, AI, Cryptocurrency, Internet of Things, Social networks, Decentralized networks, etc. Everything in life, however trivial or hard, is a nail that can be solved with those fancy hammers. So we can have a tablet juxtaposed on a fridge, a bluetooth speaker integrated to a salt shaker, a subscription model for juicers, and machine learning for all kinds of social problems. I'm frustrated by this as well. I see these days, especially, junior engineers excited about machine learning analyzing data sets that could be tackled easily with learning the problem space and excel or a jupyter notebook. 90% of the work in analysis problem is, usually, identifying data sources and possibly curating the records to a workable data set. So it really doesn't help much to apply ML off the bat. If normal regression and human inspection don't help, maybe then you can iterate on the problem and try some ML approaches. So why are junior engineers trained to think this way? Are their professors and guest lecturers excited about these approaches and teaching the students about their favorite hammers?

ghostcluster 8y ago
I'm not so sure when it comes to machine learning. Those techniquss represent such a fundamentally different approach to pattern recognition that it enables the tackling of whole classes of problems that couldn't previously be sufficiently automated with hand-coded heuristics and straightforward statistics.
- johannes1234321 8y ago
  True. But many many many problems can be solved without that hammer as well.
- omginternets 8y ago
  This is undeniably true, but consider that 99% of classification problems encountered in the industry can be solved with a logistic (if not linear) regression.
- gaius 8y ago
  Those techniquss represent such a fundamentally different approach to pattern recognition You are kidding right? With the exception of a few modern breakthroughs like GAN the vast majority of ML is 70s and 80s ideas that Moore’s Law has just taken mainstream.
  - omginternets 8y ago
    The fact that these approaches are now tractable represents a paradigm-shift in what we can do with computers. That's what the parent-poster meant.
  - ghostcluster 8y ago
    70s and 80s ideas that couldn't be implemented at scale due to the lack of massively parallelized GPU architectures with trillions of transistors to simulate the neurel net layers. It was only in the mid-late 90s that simple handwriting / character recognition techniques were able to be used in somewhat real-time.
  - sanxiyn 8y ago
    > With the exception of a few modern breakthroughs like GAN And ReLU (2011). And Glorot initialzation (2010). And He initialization (2015). And Dropout (2014). And Batch Normalization (2015). And Adam optimizer (2014). And distillation (2015). And... None of these were known or used in 70s and 80s and performance difference is enormous.
pjmlp 8y ago
The biggest issue is that it is only part of the problem, the other side is that it also infects HR requirements and now everyone needs to be a highly skilled motived engineered with 5+ years experience and several deployed applications to go through the HR filter and interviews driven by those junior engineers.
qntty 8y ago
So they can put on their resume that they have experience with machine learning? I don't have any shame about overdoing things when I think it will improve my resume.
reaperducer 8y ago
>I see these days, especially, junior engineers excited about machine learning analyzing data sets that could be tackled easily with learning the problem space and excel or a jupyter notebook. Ditto. I watched the entire Google I/O Stevenote, and was surprised how many of the "problems" it's tackling with "AI" have already been solved by other, simpler solutions. To be sure, there were some impressive moments. But much of it was reinventing the wheel with an AI spin.
currymj 8y ago
they want to get well-paid, high prestige jobs using machine learning. whereas doing the exact same task with a regression in Excel turns it into low-prestige clerical work. i think they’re acting pretty rationally once you consider their own incentives. (plus of course, the general cultural perception that machine learning is cool, smart, and cutting edge must rub off on them too and how they perceive themselves.)
- humanrebar 8y ago
  > i think they’re acting pretty rationally once you consider their own incentives Is that even true? Or are candidates and companies just caught up in a cycle of signaling? In my experience, there is amazing demand for pretty straightforward software engineering, and the talented leaders there are coveted and paid very well. Analytics, ML, etc. are desired as well, but are a minority portion of the demand.
  - HumanDrivenDev 8y ago
    If you're in a '2nd tier' first world country like Canada, Australia, the UK, Germany etc, Machine Leanring is one of the only ways you can get a salary comparable to a 'normal' Software Dev in the US. Keep that in mind.
    - kamikaz1k 8y ago
      Which countries are first tier?
      - HumanDrivenDev 8y ago
        Not many, Switzerland and Norway are all that come to mind. The real point I am trying to get across is the huge gulf between the career of a US software engineer vs almost anywhere else in the world.
- goatlover 8y ago
  I thought the well-paid, high prestige jobs required a Phd in the field? Or has machine learning become widespread enough for the unwashed programming masses?
trisimix 8y ago
As a student in CS I see myself falling for this simply because other hammers have learning curves.
nerdponx 8y ago
It's an inverted approach to problem-solving that is fine for side projects and self study but inappropriate for getting things done. I suspect part of the problem is that people don't really have enough time for self study, so they, consciously or subconsciously, start finding excuses to inject their interests into their work. I know I am guilty of having done this in the past.
SlowRobotAhead 8y ago
Definitely applies to all of that and blockchain if you want to separate that from cryptocurrency. “We know what it is - just not what it’s for... yet”
- netsharc 8y ago
  The other day I read about a company wanting to prevent ticket touting by using... blockchains. Erm, what... the... fuck? The hype around cryptocurrencies is that you can use it without trusting a single entity (a central bank). Only 1 entity determines whether an event ticket is worth anything or not, and that's the ticket checker at the gate. Why the eff would you need blockchain for this? But anyway, the startup probably got some millions of dollars, they're probably smart because they can see how stupid investors are...
  - SlowRobotAhead 8y ago
    I’ve just been replacing blockchain when reading with “mathy database” and yea a lot of context doesn’t hold up.
staticassertion 8y ago
Because technology is more than practical, it's also cool.
bassman9000 8y ago
So why are junior engineers trained to think this way? For us, it can be a competitive advantage, as thousands of engineers dismiss more traditional techniques in favor of sub-optimal, but fancier ones. And, a few years from now, it will be another powerful tool, as the The Next Big Thing takes over.
- TeMPOraL 8y ago
  As long as we find other ways to get through HR filters.
kriro 8y ago
I'll play devils advocate and disagree. I'll claim that deep learning is the greatest hammer I have seen in my lifetime and it can tackle all nails you throw at it (universal approximation theorem). Additionally it is easy to apply and get very reasonable results. At least for me it's easier to whip up a Jupyter Notebook and use some basic heuristics and a library like fastai than to struggle through Excel. Assuming that the data is available and curated (I'd argue this is a hard and very value creating task).
mwfunk 8y ago
I think these are just the normal hype cycles for new technologies. Some technology (maybe new, maybe not) captures the imagination of the industry, and there's a flurry of mostly silly activity where people are trying to figure out what else that technology can be used for. It feels like there's low hanging fruit waiting to be plucked, but no one knows what's actually going to stick. Eventually the hype dies down as a consensus forms on what it is and isn't good for. From a money perspective, there's also a gold rush mentality. VCs are going to be drawn to these sorts of things, even knowing that 99% of it will turn out to be a square peg in a round hole, because somewhere in there, there might be a unicorn. People seeking investment from VCs might also promote the buzzword of the day for their own marketing purposes (to investors, the public, and the press).
- reaperducer 8y ago
  >I think these are just the normal hype cycles for new technologies. You're not wrong. Remember in the late 90's when everything was all i-this, and cyber-that? Everything was going to be on "the interweb!" Those big mainframes of yesteryear? We don't need them anymore. We have AOL and thin clients! Back in the 80's: We don't need rockets anymore, we have space shuttles! In the 70's: Who needs stereo? We have quadraphonic sound now! Look at old mainstream magazines from the 1950's. EVERYTHING was going to be solved by being "atomic." Too busy to cook dinner? Let your atomic kitchen do the work for you! Spilled coffee on your tie? Let an atomic washer make it clean for you! Little Johnny put in the dunce corner again? Slap an atomic Think-o-Tron on his head and let radiation smarten him up! I'm sure there are a brazillion other/better examples. Feel free to add yours.
scottLobster 8y ago
It's Anecdata, but when I was coming up through my Computer Engineering program a few years ago there were roughly three categories of students: 1. The Traditionalists: Students who are fascinated by the technology and what it can do at its limits. These people are disproportionately drawn towards the more arcane/unsexy applications (RF, hardware, high-power, radars, antenna design, IC design, embedded systems, robots, rockets, quantum computing, etc) and research. 2. The New Wave: Students who were inspired by a Steve Jobs keynote, these people were almost exclusively interested in potentially mass-market user-facing applications and have a greater emphasis on image (both personal and professional). 3. The Unmotivated: People who are there "because it pays well" and for no other reason. Not many of them made it to graduation, and the ones that did were "meh" engineers at best. Most of the naive junior engineers you're seeing probably belong to the modern incarnation of category 2. Steve Jobs has been dead long enough for new grads to have been inspired by the next generation of buzzwords. They're fine engineers in general, but they have a sometimes unhealthy fixation on said buzzwords and how "future" they are, because that's what got them into engineering. Most will concede when a simpler solution is presented, but it's uncomfortable for them. Unsexy but effective techniques undermine their aesthetic impulses and, to a degree, their self-image as engineers. A good popular example is Elon Musk refusing to use LIDAR on Teslas, and introducing way too much automation, too fast at Tesla's factory. People, even junior engineers who think they're part of a "ground-breaking" movement often have trouble separating "obsolete" from "tried-and-true".
- palisade 8y ago
  4. The Savant: Also known as the absent minded professor. You could throw questions at this person in an interview and they will seem completely inadequate and unprepared. But, then apply them to a problem on the job and like a juggernaut they will defeat any obstacle. They will even solve problems your best engineers have been struggling with for years without breaking a sweat. A lot of companies make the mistake of weeding this type of engineer out in the interview process.
  - abhishekjha 8y ago
    I may qualify in this category. Companies reject me because I gave them a brute force solution for finding repeated values in an array. But there hasn't been any single instance with my current job that I couldn't check all of the requirements that were assigned to me, even those which I thought I wasn't prepared for. The hiring/evaluating process needs to get refined.

ghostcluster 8y ago

I'm not so sure when it comes to machine learning. Those techniquss represent such a fundamentally different approach to pattern recognition that it enables the tackling of whole classes of problems that couldn't previously be sufficiently automated with hand-coded heuristics and straightforward statistics.

johannes1234321 8y ago
True. But many many many problems can be solved without that hammer as well.
omginternets 8y ago
This is undeniably true, but consider that 99% of classification problems encountered in the industry can be solved with a logistic (if not linear) regression.
gaius 8y ago
Those techniquss represent such a fundamentally different approach to pattern recognition You are kidding right? With the exception of a few modern breakthroughs like GAN the vast majority of ML is 70s and 80s ideas that Moore’s Law has just taken mainstream.
- omginternets 8y ago
  The fact that these approaches are now tractable represents a paradigm-shift in what we can do with computers. That's what the parent-poster meant.
- ghostcluster 8y ago
  70s and 80s ideas that couldn't be implemented at scale due to the lack of massively parallelized GPU architectures with trillions of transistors to simulate the neurel net layers. It was only in the mid-late 90s that simple handwriting / character recognition techniques were able to be used in somewhat real-time.
- sanxiyn 8y ago
  > With the exception of a few modern breakthroughs like GAN And ReLU (2011). And Glorot initialzation (2010). And He initialization (2015). And Dropout (2014). And Batch Normalization (2015). And Adam optimizer (2014). And distillation (2015). And... None of these were known or used in 70s and 80s and performance difference is enormous.

pjmlp 8y ago

The biggest issue is that it is only part of the problem, the other side is that it also infects HR requirements and now everyone needs to be a highly skilled motived engineered with 5+ years experience and several deployed applications to go through the HR filter and interviews driven by those junior engineers.

qntty 8y ago

So they can put on their resume that they have experience with machine learning? I don't have any shame about overdoing things when I think it will improve my resume.

reaperducer 8y ago

>I see these days, especially, junior engineers excited about machine learning analyzing data sets that could be tackled easily with learning the problem space and excel or a jupyter notebook. Ditto. I watched the entire Google I/O Stevenote, and was surprised how many of the "problems" it's tackling with "AI" have already been solved by other, simpler solutions. To be sure, there were some impressive moments. But much of it was reinventing the wheel with an AI spin.

currymj 8y ago

they want to get well-paid, high prestige jobs using machine learning. whereas doing the exact same task with a regression in Excel turns it into low-prestige clerical work. i think they’re acting pretty rationally once you consider their own incentives. (plus of course, the general cultural perception that machine learning is cool, smart, and cutting edge must rub off on them too and how they perceive themselves.)

humanrebar 8y ago
> i think they’re acting pretty rationally once you consider their own incentives Is that even true? Or are candidates and companies just caught up in a cycle of signaling? In my experience, there is amazing demand for pretty straightforward software engineering, and the talented leaders there are coveted and paid very well. Analytics, ML, etc. are desired as well, but are a minority portion of the demand.
- HumanDrivenDev 8y ago
  If you're in a '2nd tier' first world country like Canada, Australia, the UK, Germany etc, Machine Leanring is one of the only ways you can get a salary comparable to a 'normal' Software Dev in the US. Keep that in mind.
  - kamikaz1k 8y ago
    Which countries are first tier?
    - HumanDrivenDev 8y ago
      Not many, Switzerland and Norway are all that come to mind. The real point I am trying to get across is the huge gulf between the career of a US software engineer vs almost anywhere else in the world.
goatlover 8y ago
I thought the well-paid, high prestige jobs required a Phd in the field? Or has machine learning become widespread enough for the unwashed programming masses?

trisimix 8y ago

As a student in CS I see myself falling for this simply because other hammers have learning curves.

nerdponx 8y ago

It's an inverted approach to problem-solving that is fine for side projects and self study but inappropriate for getting things done. I suspect part of the problem is that people don't really have enough time for self study, so they, consciously or subconsciously, start finding excuses to inject their interests into their work. I know I am guilty of having done this in the past.

SlowRobotAhead 8y ago

Definitely applies to all of that and blockchain if you want to separate that from cryptocurrency. “We know what it is - just not what it’s for... yet”

netsharc 8y ago
The other day I read about a company wanting to prevent ticket touting by using... blockchains. Erm, what... the... fuck? The hype around cryptocurrencies is that you can use it without trusting a single entity (a central bank). Only 1 entity determines whether an event ticket is worth anything or not, and that's the ticket checker at the gate. Why the eff would you need blockchain for this? But anyway, the startup probably got some millions of dollars, they're probably smart because they can see how stupid investors are...
- SlowRobotAhead 8y ago
  I’ve just been replacing blockchain when reading with “mathy database” and yea a lot of context doesn’t hold up.

staticassertion 8y ago

Because technology is more than practical, it's also cool.

bassman9000 8y ago

So why are junior engineers trained to think this way? For us, it can be a competitive advantage, as thousands of engineers dismiss more traditional techniques in favor of sub-optimal, but fancier ones. And, a few years from now, it will be another powerful tool, as the The Next Big Thing takes over.

TeMPOraL 8y ago
As long as we find other ways to get through HR filters.

kriro 8y ago

I'll play devils advocate and disagree. I'll claim that deep learning is the greatest hammer I have seen in my lifetime and it can tackle all nails you throw at it (universal approximation theorem). Additionally it is easy to apply and get very reasonable results. At least for me it's easier to whip up a Jupyter Notebook and use some basic heuristics and a library like fastai than to struggle through Excel. Assuming that the data is available and curated (I'd argue this is a hard and very value creating task).

mwfunk 8y ago

I think these are just the normal hype cycles for new technologies. Some technology (maybe new, maybe not) captures the imagination of the industry, and there's a flurry of mostly silly activity where people are trying to figure out what else that technology can be used for. It feels like there's low hanging fruit waiting to be plucked, but no one knows what's actually going to stick. Eventually the hype dies down as a consensus forms on what it is and isn't good for. From a money perspective, there's also a gold rush mentality. VCs are going to be drawn to these sorts of things, even knowing that 99% of it will turn out to be a square peg in a round hole, because somewhere in there, there might be a unicorn. People seeking investment from VCs might also promote the buzzword of the day for their own marketing purposes (to investors, the public, and the press).

reaperducer 8y ago
>I think these are just the normal hype cycles for new technologies. You're not wrong. Remember in the late 90's when everything was all i-this, and cyber-that? Everything was going to be on "the interweb!" Those big mainframes of yesteryear? We don't need them anymore. We have AOL and thin clients! Back in the 80's: We don't need rockets anymore, we have space shuttles! In the 70's: Who needs stereo? We have quadraphonic sound now! Look at old mainstream magazines from the 1950's. EVERYTHING was going to be solved by being "atomic." Too busy to cook dinner? Let your atomic kitchen do the work for you! Spilled coffee on your tie? Let an atomic washer make it clean for you! Little Johnny put in the dunce corner again? Slap an atomic Think-o-Tron on his head and let radiation smarten him up! I'm sure there are a brazillion other/better examples. Feel free to add yours.

scottLobster 8y ago

It's Anecdata, but when I was coming up through my Computer Engineering program a few years ago there were roughly three categories of students: 1. The Traditionalists: Students who are fascinated by the technology and what it can do at its limits. These people are disproportionately drawn towards the more arcane/unsexy applications (RF, hardware, high-power, radars, antenna design, IC design, embedded systems, robots, rockets, quantum computing, etc) and research. 2. The New Wave: Students who were inspired by a Steve Jobs keynote, these people were almost exclusively interested in potentially mass-market user-facing applications and have a greater emphasis on image (both personal and professional). 3. The Unmotivated: People who are there "because it pays well" and for no other reason. Not many of them made it to graduation, and the ones that did were "meh" engineers at best. Most of the naive junior engineers you're seeing probably belong to the modern incarnation of category 2. Steve Jobs has been dead long enough for new grads to have been inspired by the next generation of buzzwords. They're fine engineers in general, but they have a sometimes unhealthy fixation on said buzzwords and how "future" they are, because that's what got them into engineering. Most will concede when a simpler solution is presented, but it's uncomfortable for them. Unsexy but effective techniques undermine their aesthetic impulses and, to a degree, their self-image as engineers. A good popular example is Elon Musk refusing to use LIDAR on Teslas, and introducing way too much automation, too fast at Tesla's factory. People, even junior engineers who think they're part of a "ground-breaking" movement often have trouble separating "obsolete" from "tried-and-true".

palisade 8y ago
4. The Savant: Also known as the absent minded professor. You could throw questions at this person in an interview and they will seem completely inadequate and unprepared. But, then apply them to a problem on the job and like a juggernaut they will defeat any obstacle. They will even solve problems your best engineers have been struggling with for years without breaking a sweat. A lot of companies make the mistake of weeding this type of engineer out in the interview process.
- abhishekjha 8y ago
  I may qualify in this category. Companies reject me because I gave them a brute force solution for finding repeated values in an array. But there hasn't been any single instance with my current job that I couldn't check all of the requirements that were assigned to me, even those which I thought I wasn't prepared for. The hiring/evaluating process needs to get refined.