Hacker News new | past | comments | ask | show | jobs | submit login
Humanity and Machine Learning [video] (a16z.com)
46 points by jadk157 7 months ago | hide | past | web | favorite | 10 comments



I gave a lecture at the University of Toronto last year with a similar point to what the speaker was getting at. Namely, we should be focusing on augmenting human intelligence, not chasing some sort of future robot overlords. See https://adamdrake.com/novel-results-considered-harmful.html

This was the approach Douglas Englebart took in the lead up to the Mother of All Demos (MOAD) and it's unfortunate people have forgotten that bit of history.

There is so much that can be done to improve the lives of people if we simply focus on how to optimize a variety of things that have to happen every day. This doesn't mean we should forsake longer-term goals like fully autonomous vehicles or other things, but it does mean that some of those who are most capable of making a very real and significant impact in the lives of thousands of people are forsaking that opportunity.

There is also often the argument that such near-term thinking will lead us to a local optimum in terms of technology advancement, but I don't see that there has been significant evidence to support that claim. After all, Englebart's demo of version control and collaborative editing (among other things) was over 50 years ago and that didn't seem to stall technological progress.


> we should be focusing on augmenting human intelligence

Who’s we? I can throw machine learning at practical real world problems right now.

Augmenting human intelligence? I’m not aware of any work in this area.


Great! We (technologists) should be using machine learning approaches to practical real world problems, particularly those problems whose solutions allow humans to be more productive. That's what I mean by augmenting human intelligence/capabilities.


Would someone be willing to provide a brief summary for those who cannot watch the video at the moment?


The video walks through interesting uses of ML in industry (mostly outside normal FAANG applications like optimizing your news feed) e.g - precision farming - blood delivery using drones - monitoring construction sites - writing more diversity-friendly job descriptions - collecting relevant legal documents

Speaker also argues that we should focus on leveraging ML-human collaboration to surpass human performance, instead of falling into a "robots will replace us" narrative


> Speaker also argues that we should focus on leveraging ML-human collaboration to surpass human performance, instead of falling into a "robots will replace us" narrative

I didn't watch the full video, so I am only responding to your summary. These kinds of sentiments, where technologists profess that technology will always make everything OK, while not at all addressing how our current economic system will fail large swaths of our society if automation comes to fruition (beyond "new jobs will come up!" or "ignore the luddites!") strikes me entirely as a "Let them eat cake" attitude.

Here's a thought experiment: What would happen if, by the end of 2019, true, 100% self-driving cars became a reality. I know this is not going to happen, but it no longer seems a far-fetched fantasy. In the US, driving is the number one job for the majority of states. What are all of these people supposed to do, become self-driving car programmers?

I am a big believer in technology but I am very worried for the future of society.


> Here's a thought experiment: What would happen if, by the end of 2019, true, 100% self-driving cars became a reality. I know this is not going to happen, but it no longer seems a far-fetched fantasy. In the US, driving is the number one job for the majority of states. What are all of these people supposed to do, become self-driving car programmers?

In the "let them eat cake" scenario, those people will have plenty of free time to stage a revolution and take their piece of the cake by force. Below the threshold of civil war, you get a Wild West where armed bandits hold up trains of self-driving trucks and loot them, creating a thriving job market for security guards protecting the trucks.

Of course both of those scenarios are highly undesirable, so it's more likely that some kind of tax will be introduced to take automating jobs from extremely profitable to barely profitable, with the proceeds used to pay for unemployment benefits and retraining for the displaced workers. (With the amount of retraining depending on how many other jobs haven't been automated yet.)

That said, some people are going to see their standard of living decrease without any way to escape. Such people have existed since forever, and they usually end up homeless if not dead. Automation doesn't create any new problems in that regard, it just makes them large enough that they can no longer be ignored easily.


If i were to play the devil's advocate, these "Skilled" jobs have always been on the line. Look at what Uber and Lyft have done to F.T. cab drivers in all major U.S. cities. I have been using Scoop(Bay area wide Corporate Car share app). Its only a matter of time FAANG dumps their "Google bus" in favor of their employees scooping to work by incentivizing and subsidizing it. So there goes the cushy "Google bus" driver job. Not by automation but by a variation of mass transit. In such cases, the employers have to pick up the tab by offering those "skilled" workers displaced OJT/retraining and funneling them into more "non-skilled", knowledge-based jobs. It is not going to be easy and Corporate America finally has to acknowledge its Corporate responsibility. If not Mayhem will ensue and everyone stands to lose.


The video makes the argument that long haul truck driving is the job killing people most effectively. Besides fatal accidents the lack of physical activity is reducing life span. 69% type 2 obesity, 17% type 3 obesity and 27% get < 6 hour sleep. He also mentions 50 000 driver shortage


Funny. You are advocating killing the reason that made driving the major job provider.

I can only imagine what you’d write at the time when first cars came to be or combustion engine ...

Your position is evidence for the power of propaganda.




Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: