Hacker News Reader

Well, that there are 5000 employees suggests we don't have to worry quite yet.

I'll panic when the headline is "Microsoft Forms New Zero Person AI Division" :)

Or even worse... "Cortana Forms New AI Division at Microsoft."

I'll do you one better: "Tay Forms New AI Division at Microsoft."

Tay hires a new human relations manager

Headlines are for us - will we necessarily know when we have been superceeded.

I think it's important to understand that is really happening here. Microsoft is merging Microsoft Research with their application oriented AI work. See the press release from Microsoft (http://mnc.ms/2ddF5dC).

MSR is already a pretty big group at Microsoft and does all sorts of work that is completely unrelated to AI. Now, I am confident the teams within MSR will have to start thinking about AI, and how their work relates to AI, but this initiative is much more of an interdisciplinary effort than a 5000 person deep learning team.

Exactly. Not a single mention of Deep Learning.

Well, companies are certainly spending a lot of money on something that they are branding AI - I might wait and see how much of it is on real AI - what ever that is.

While most people in the industry were busy debating the usefulness and definition of "cloud", Amazon and others were busy building it. To me this is similar to that situation. There is real progress happening with practical benefit in AI, machine learning, deep learning, and Microsoft is committed to being at the forefront. Personally I'm not waiting. I'm teaching myself data modeling and other techniques that'll let me be part of the coming revolution.

They don't have to achieve "real AI"; they only need to come up with things that do more work for us to win.

Real AI is a fool's goal. Once we achieve that, we will simply be obsolete, I'm in no hurry to get there.

haha yeah in this case, we want to shoot for the moon but definitely want to hit the tree

Prefer the terms "strong AI" or Artificial General Intelligence.

>They don't have to achieve "real AI"; they only need to come up with things that do more work for us to win.

Yeah, can't wait for the new Clippy.

Great point. It seems to me that people are categorizing anything that involves analytics as AI now:

"I asked Cortana to query our clickstream data and she said that men over the age of 50 like to click on Viagra ads."

This is not AI. Although there is some real research going on, it seems that everyone is jumping on the bandwagon and using AI as a buzzword.

AI will certainly be a feature in all software solutions by 2020.

The warnings of "Second AI Winter" are in response to the glut start-ups whose sole products are ML/AI platforms.

It's wiser to invest in real solutions that happen to use AI models.

I think 200k as an average is good, since some will be making a whole lot more, and there will be the 100K per year positions too.

My equivalent to your dearth of people in the know in 2011, was studying AL (Artificial Life) in 1990s, and nobody heard of it. It included the study of ANNs, GAs, GP and AI in general (which I prefer to call CI - Computational Intelligence nowadays). The book that started it all for me [0].

There were no immediate applications aside from expert systems here and there, and the fuzzy logic appliance controllers coming out of Japan. However, now, with self-driving cars, recommender systems, image recognition (face-matching surveillance post 9/11) have big pay-offs or budgets to spend on it.

VR is having its second renaissance. I thought the first time it would have taken off even with the clunky glasses and headsets, since gaming was already such a huge money industry.

And now with modeling becoming prominent again, AL paradigms are being modified, created and repurposed for all sorts of cool things. I play with NetLogo since it is a fun environment for that [1].

My only regret is that I was in at the beginning, but left it to pursue other things, and so I am not at the level of practice to get one of those 200K jobs. I still kept studying it all these years though, and I have coded my own bits and pieces, but mainly for conceptual pieces, art and music, not practical applications. I play with Darknet [2] now, since C was my second language after Assembly (6502 and then x86_32), and it is great fun, and fast. A very understandable and manageable platform.

I am hoping it all leads someday to keeping me alive longer to enjoy studying some more, because as I get older that's really what I enjoy most aside from family!

[0] http://www.springer.com/gp/book/9780387976143

[1] https://ccl.northwestern.edu/netlogo/

[2] http://pjreddie.com/darknet/

Well its not 5000 new people I assume ~70% would be restructuring (E.g. as the article states from Bing and Cortana). Even then at 1500 new employees, its a huge undertaking and vote of confidence into importance of AI.

Except for severance and hiring bonuses, what's the difference between restructuring someone's job, or firing them and hiring another person? In the end, you're still committing $1B/year to AI jobs.

Sure, its a ton of money, but they've had $20+ billion/quarter in revenue the last four quarters.

https://finance.yahoo.com/quote/MSFT/financials?p=MSFT

Well a portion of $1B is already accounted for in the balance sheet and budget.

>(which is conservative if you include benefits, etc.)

I've never been able to figure out how my employer says I cost them $280 per hour.

If you're working on a service organization, your rate is defined by:

Billable hour = salary cost + benefits + G&A + overhead

Where:

- Salary cost = gross salary / 1800 (or whatever # hours/yr your company uses)

- Benefits = proportional cost of vacation, 401K, etc

- G&A = General expenses and admin (office, equipment, free food, etc)

- Overhead = amount needed to cover everyone else who isn't billable (HR, Finance, IT, assistants, President, etc)

If you do the math, you'll see $200-300 is a standard rate in high cost markets (NYC, SFO) for someone with a six-digit figure. More junior positions are ~$150, while very senior roles (Partner, Director, etc) would be in the $300-500 range.

Because your boss has no billable hours. Nor his boss, nor the boss above him. The front desk guy, the lady in accounting, the payroll people, facilities, most of the IT folk.

You need one of each these people for every N developers (different value of N for different classes of service). Therefore the cost of having you includes the 1/N cost of each of these people.

Maybe it is because you are burning $20 bills every 5 minutes at work!? They have been trying to make you stop for a long time.

Are the stock options involved based on unicorn valuations?

It's that extra piece of coal you insist on burning.

$280/hour is $280 * 40 * 52 = 582400/year btw. I think your figures are off.

I need to bill out at an amount higher that that to be profitable to the company. They don't pay me anything near that.

Depends what you do.

If you are client-facing consultant, your billable rate is paying:

- Your salary

- Your benefits

- Your expenses (travel, per diem, equipment, software, anything else)

- And then... Salary, benefits, expenses of all your NON-client-facing teammates: Salespeople, marketing, management, client relationship, administrative, HR, etc.

So as a consultant, you are the product, and as such have to be sold at a price that nets enough profit to pay for rest of the organization.

Normally when they say "you cost $150/hour", what they mean is, "you, and all the support framework behind you, needs to be billed at $150/hour to pay off".

Hope that helps. If you're in-house developer and not client-facing, it's harder to understand and justify cost figures, but typically would be nowhere near that rate...

Are you burning $50s?

Or maybe just really, really bad at your job?

If you look at MS's financial statements you can (almost) do the math. Divide R&D spending by R&D personnel and you end up at $300k, which can serve as an upper bound. That supposedly includes vendor budget (not included in #personnel), of which MS employs quite a few, and overhead. If you do the same with Google you end up at $500k, which seems insane.

Where did you get that this is entirely Deep Learning?

Not a single mention of that anywhere.

Also, am I misunderstanding, or is the "Deep" part of Deep Learning not a very quantifiable thing? I.e. there's no rule that says "Your neural net isn't 'deep' unless it has X amount of layers or N amount of nodes per layer" or anything. It just specifies using large neural networks for examining a huge amount of data and feature set. Right?

So really, any AI-based learning is going to be "deep" from now on, simply because we've reached the point where we can handle large neural nets and complex datasets, right?

Microsoft is spending a ~billion dollars a year on their entire research department, which AI is a part of.

>... there will be no "AI industry"; it will branch out, and there are more potential applications of "understanding data and automating decisions" than there were in the 80s.

IBM was the first company with a massive investment in AI (during this cycle) - they are now trying to push Watson as a product. Did it pay off so far? I am not quite sure.

>Pretty amazing when I think back to 2011 and there were machine learning conferences where literally no one I spoke to had heard of deep learning.

You'll and most people too be replaced with a machine and have no job, pretty amazing.

Can you help us understand what you think this new division will be producing? What are a few deep learning-enabled products or projects that are in production and generating revenues?

>but this time it's different

Don't they always say that?

Yes, they do. Sometimes they're wrong, but sometimes they're right!

But this time is different, so this time they are right!

"$200k per year" is only conservative if you look at tech jobs in San Francisco.

It is an substantial amount of money in the rest of the world.

Microsoft historically pays slightly better than Google and other valley companies. And, the cost of an employee is usually significantly higher than the salary alone. Something like 30%-60% higher, when factoring in benefits, cost of providing offices, stock options, etc. So, $200k is pretty conservative in many tech hubs, not just the valley, particularly when you're talking about people involved in research (who are more likely to have obtained graduate degrees).

Where are you getting this information from? I got an offer from Microsoft and it definitely wasn't even close to what people who get offers from FB or Google get.

Anecdotally, pay I've seen from Google and FB are higher than Microsoft/Amazon, but of course that's not saying much. Still, the first source I can find matches what I've seen: https://blog.step.com/2016/04/08/an-open-source-project-for-...

You realize you're comparing job offers in the most expensive city in the western hemisphere to offers in Seattle, right?

I was only strictly considering pay because that's what the post/comment thread was about.

Also, Google and Facebook have offices in Seattle, and from what I understand the pay between their HQ and Seattle offices are the same.

it's worth more in seattle because there is no state income tax!

Soon to be the second most expensive city.

My information may be out of date. My friends who've chosen between Google and Microsoft got better offers from MS; though often ended up going to Google anyway, because they just liked it better. But, it was all several years ago.

Google has a consistent reputation for paying a bit lower than other giants.

Microsoft pays well, but this is the first time I've heard them paying better than google. When I was at MS, there were always complaints about losing potential hires to Google over much higher salaries.

Comparing MS salaries in Redmond to Google salaries in Mountain View is rather apples and oranges. Cost of living is different (esp. once you start shopping for houses), state taxes are drastically different etc.

Now, Google also has an office in Kirkland. That would be an interesting place to compare. But it's also not particularly large.

I don't appreciate the "cost of living" salary differential excuse, and let me tell you why... Ask yourself, does Microsoft charge me less for Windows or Office if I live in, say, Kansas? No, they don't.

Why would that matter in the slightest when you're actually shopping for a job? Either the pay is more (in terms of quality of life it translates to while residing where you need to live to commute), or not.

Relatively little of your money is spent on those kinds of things though. If you're like most people, most of your money goes to taxes, a place to live, and food. Those are location-dependent.

And free food. Can't forget about the free food.