It's interesting to talk about what "IBM" has been doing, because after all the A&M IBM has done this last decade, they've actually got plenty of "activity" to their name. Just, activity that was previously known under the names of Softlayer, Cognos, Redhat...
IBM itself is just a consulting "solution provider" company that provides "solutions" in terms of the products and services of its subsidiaries. They're never going to look like they're on the cutting edge of some space, unless that's what one of their (equally-stodgy) enterprise clients wants to pay them to do. (And even then, why build when you can buy? IMHO IBM would only ever build a new technology at this point if there was no existing company out there in the market, with that tech IP, that they could just acquire.)
Certainly, there's the Watson AI lab, though I don't think anyone's expecting much of it at this point; nor, importantly, was anyone inside IBM ever expecting much of it. Watson (the lab specifically, not the brand strategy of calling all their cloud OLAP SaaS products "Watson") is more a nod to their clients to say "of course we're looking into AI", and a pool of people competent enough that they could clone the existing ML approach of one of their competitors if one of their clients demanded it.
IBM could be a really innovative company if they rearranged their talent pool, certainly. But innovation—getting ahead of the market and so needing to educate the market on the problem their products solve—hasn't been IBM's business model since the 1950s. IBM makes money by listening to clients' "unique" needs, and then meeting them with "custom" solutions at low CapEx cost, by using an ever-larger flywheel of existing solutions pre-tuned to almost fit every possible business use-case.
I suppose its worth noting that IBM folks received a record 9,262 U.S. patents in 2019 - the most patents ever awarded to a U.S. company and the 27th consecutive year of it grabbing the no.1 patent spot.
No idea how valuable or inovative any of these patents are, but I think its important to note when discussing: "They're never going to look like they're on the cutting edge of some space"
Appearing on the cutting edge to enterprise clients is important to IBM. That doesn't mean they are on the cutting edge (they might or might not be). IBM's large pool of patents is something they can advertise to clients. Most companies don't see their patents as valuable, and thus don't patent as much stuff.
I've worked with former IBM people - all (or some it seems) of them have a plaque of whatever their first patent was. When they talk about it they admit it was a trivial invention that is too specific to be of much use outside the one project (thus there is no value in keeping it from competitors). IBM encourages patenting everything and has people looking for things to patent. Most other companies only patent things that are unique and worth the costs. (keeping from competitors, licensing to competitors, or keeping competitors from patenting it and stopping your use)
I worked at IBM Research (Yorktown Hieghts, Last century, assisting infrastructure for research).
They spent 6 Billion Dollars on research that year (late 90s)
But Its true, IBM Values patents quite heavily. You can see some of them they printed as wall paper in the lobby (Along with mechanical calculators and models of some of daVinci's machines.)
But they also have "Trade Secrets" which are things they think they're competitors won't figure out and if you don't patent it you aren't telling the world how its done. I think some of the chip chemicals and processes for working with silicon wafers were to be classified as such.
They seemed always to be pushing the researchers to make something they could sell..
As an aside, IBM Yorktown is an oddly round building..
A few years ago, one of my buddies got a research scientist job at IBM T. J. Watson Research Center at Yorktown Heights and I got a friend&family pass visiting their building there - inside it's amazing, considering this was built in 1950s. I remember there was a long hallway with invention exhibits, some of the most iconic inventions in computer hardware, ICs, CPUs were on display. Pretty amazing.
The arc-shape building reminded me Apple Park in Cupertino, except TJ Watson center was built decades earlier.
I did enjoy the building. Rather a lot of interesting stuff was thought up there and it was humbling.
I like that the hallway was on the outside and hallways radiated out (and everyone had offices with frosted glass). It was a pain running pipes though as standard bend didn't quite work for that radius.
I'm really waiting for the day that software patents are finally struck down and we can stop this farce of pretending software patents are actually useful and valuable beyond being a cudgel in legal battles.
I just finished (a light) editing of an English (machine translated) version of a book for someone I have a lot of respect for who is a Japan-based mentor of mine - Hiroshi Maruyama had 26 years in IBM research, eventually as the director of IBM Tokyo Research Labs.
Some of the key points of the IP marketplace were
- fees for the licensee fee would be set upfront;
- the patentee would pay a % value to the government as a tax;
- if the patentee wanted to use the patent exclusively, a high licensing fee could be set, but that would also result in a higher tax being due to be paid.
The approach was to dis-incentivize bad actors putting monopolies on products via patents. It would also encourage higher utility value from patents under this system.
Another interesting concept Maruyama's book covered was LOT networks used to fight against Trolls.
- https://lotnet.com/
The idea with the Lot network was that if a patent did get rights to a patent from a member of the lot network, If any of the patents from members falls into the hands of a patent troll, all other members are automatically granted a free license of the patent.
The book covers themes on how to do research in corporate organizations (the audience focus remains with Japanese researchers/students) and has some interesting discussions.
I will provide a link to the book when it is available (if there is interest).
Maybe, it certainly sounds much better at a glance than the current system though there's tough stuff like how to set that value in such a way that independent inventors aren't forced to partner with a large company in order to use the patent system. I think some reforms like that combined with a much stricter definition of "patent-able" for some categories and/or better challenge mechanisms. Getting a bad patent thrown out due to prior art is still quite expensive and trying to evaluate if a patent is non-obvious is quite difficult to litigate.
Being a cudgel in legal battles is quite a lot of value though, that's basically the point and is how patents are meant to support and reward innovation.
There's a second part to the idea of patents though they protect innovations with the idea that once they expire other people can implement them and software patents fail abjectly at that second. There's no meat the the implementation, all modern software patents seem to boil down to connect a client to a server and display the results on an interface. There were some genuinely useful patents back in the 80s it looks like for stuff like MP3 encoding.
Also there should be some additional value to people not just companies for the government to be granting legal monopolies, and it should also be restricted to things that are actually implementable not extremely vague system level diagrams.
From what I understand, those sorts of vague software patents are far far harder to get today - in theory they never should have been granted. Unfortunately, mistakes were made, but in recent years the legal system is catching up to that, and hopefully as those older patents expire newer patents along those lines won't be granted
"On June 19, 2014 the United States Supreme Court ruled in Alice Corp. v. CLS Bank International that 'merely requiring generic computer implementation fails to transform [an] abstract idea into a patent-eligible invention'"
Software patents end up being so vague that basically every large company is violating some in some way. They don't get in legal battles with eachother often because the resulting legal decision will be somewhat random and risks huge retaliation in counter-suits.
It's not discontinuous - that's just the theory behind patents. Patents give inventors a legal "cudgel" as the commenter before put it, which makes innovation more profitable. Patents are just a legal construct, the idea is that having a patent on an invention gives you access to legal tools to protect a temporary monopoly which makes innovation more profitable, so you don't have to worry about investing money in innovation and then losing all of your profits because of copycats. However well that actually works and whether or not it ends up being a net positive for innovation is a different question, but all a patent is is a legal tool to stop others from using or making money off of your innovation, as enforced through lawsuits in the court system.
I'm not knocking them for their materials or process patents for some of their remaining R&D arms but the majority probably are software patents which I think are 95-99% unusable garbage outside of a legal fight. Generally software patents are too broad and don't contain enough implementation detail to actually use them once they do expire which is one of the important parts of the bargain for patent protection initially.
My guess is that these companies are a lot more aggressive internally about getting their employees to author patent submissions. They either actively screen code submissions and projects for patentable ideas, make patent authorship a condition of advancement, or otherwise tie authorship to performance reviews in a meaningful way.
Companies I've worked for have teams that facilitate employee patent submissions, and they will tell you why patenting is important to the company. They have had some kind of bonus structure so that you get rewarded if you submit a successful patent. But they pay well enough otherwise that, unless you're personally motivated to see your name attached to patents, it's just not worth the effort.
To put this another way: the patents you author as an employee of a company are probably not that valuable to you. But, because of the potential legal protections they offer, they are very valuable to your employer. I think the companies that are more successful in having high patent counts are doing more to align employee interest with corporate interest w/r/t patents.
There is a continuum. IBM patents everything. The other companies find patents less useful (for whatever reason) and so they are more careful about what they will spend money patenting. I've made several "inventions" that I'm sure any of the companies on your list would have patented, but since I don't work for them I don't have any patents.
As an ex-IBMer, I can tell you they really incentivize patents. IIRC, you get $1000+ on your first patent and pretty decent money on subsequent patents. Many employees form 4-5 person teams where each member basically work on their own patent but their name is listed on all patent applications submitted by their team. This way even if 4 out of 5 patents are rejected, team still has 1 approved patent and everyone in team wins some money.
If you listen to Lin Sun interview on Google's Kubernetes Podcast you hear that she have the title Master Inventor. Since she is interviewed under the title of Istio I was sure she invented some interesting techincal stuff in that field. You don't have to wait too long for (at least what I find as) the disappointing explanation of what does the title mean, why it exists and what the results are.
In short, there's incentive as well as assistance in filling up patents, they don't seem to discriminate on basis of usability, feasibility or anything else. They do give bigger bonuses for patents that are in one of the fields considered strategic for the company.
Listening to it was somewhere between amusing to depressing:
The CPUs that handle 86% of all credit card transactions for example (in the various mainframes IBM builds). Pretty nice modern architecture still capable of executing 60 year old code.
You might buy a POWER machine from a boutique manufacturer other than IBM but there were other major players such as Motorola, Apple, then later Nintendo: Nintendo adopted with the GameCube, then Wii, but IBM made custom chips for the XBox360 and the PS3, Nintendo still uses POWER in the Wii U.
I don't know why people overlook the progress IBM seems to have made in the quantum computing world. Perhaps not as impressive as Google's progress but impressive none the less.
For me personally, it was because even from the inside at IBM, you wouldn't hear much about what the quantum folks are doing if you were doing general tech. They really are in their own little world. I don't think they were even in the whole-IBM Slack team, though they might just have been security-compartmentalized with ACLs.
Was Watson just smoke and mirrors? Have there been other systems in the NLP space that eclipsed it? Genuinely asking because I don't know much background behind the story of Watson's rise and fall.
Watson didn't rise or fall, it was the centerpiece in IBM's marketing campaign, which worked for a while before they moved on. It never got beyond or matched the state of the art in anything, that wasn't the point, it got the name "Watson" on TV.
IBM wanted to prove that they employed smart people, so they hired some smart people and had them write a computer program to play Jeopardy. Did it matter that it had nothing to do with anything they were selling? Perhaps only IBM has those figures.
OK fine but there was _some_ non trivial technology behind the system that won the Jeopardy game. What I'm asking is whether what seemed like the state of the art NLP system at the time got eclipsed by newer and better systems or whether the whole thing was never really a state of the art NLP system to begin with.
The problem with Watson is that they don't have a business case. IBM sent salespeople to big customers with big problems and tried to find things to fix. In some cases, they did. But most of the time, the insurance companies, government agencies, etc they do business with scratched their heads and didn't do anything. Learning new things about their data might be seen as a threat to whatever enterprise bullshit they do!
The problem is you have companies like Google, Microsoft, Amazon, Apple, Facebook, etc have problems that this tech solves. It's easier to come up with a product with a problem that you understand. I can ask Google Photos or Siri to show me pictures of my dog in the snow in 2015, and they do. So I give Google & Apple money to store my crap. Google and Facebook use AI with all of the data they hoover up to peddle products to me. My grandparents get ads for depends, I get ads for drones, Google and Facebook make $.
Now, companies like Amazon, Microsoft and Google can go to companies that were prospected by IBM with solutions. Microsoft is minting money with ATP, because enterprise security teams suck. Amazon is selling creepy facial recognition to people, because people see it on TV, have a Ring doorbell, and want the capability. Google is selling GIS solutions, etc based on work done on maps.
> Learning new things about their data might be seen as a threat to whatever enterprise bullshit they do!
No, the problem was that Watson would be unable to help you learn anything new about enterprise data. IBM didn't even have a plausible, non-trivial proof of concept to trot out six years ago.
State of the art NLP system isn't a very meaningful term. You can't really say that Watson was "better" than Google's NLP systems at the time because they were solving different problems.
In general, what they did was not trivial, but it also wasn't revolutionary. Some of it was novel, but novel in the sense that it applied specifically to the problem they were trying to solve. It had little impact on the technology behind what IBM eventually tried to sell as Watson.
Anybody with a bucket of money could have built the same thing at the time. The impressive part was IBM figuring out that it would be worth spending a bucket of money on.
I think it was fantastic, modern NLP could do much more (if put together with the cleverness of the Watson pipeline) and it's a pot of gold if IBM ever stop trying to apply it in healthcare and apply their brains to how it could work in customer service (for example).
Note : I am aware that there are "Watson" products that claim to be this - but they aren't because they are MBA's ideas of what the best route to selling crap to the unwary. If IBM had appointed someone with a clue and given them 10x the R&D budget for the gameshow to deliver a decent product I reckon they'd have got (at least) 100x. But... oh no.. promise 30x for 1x and get f-all.x^2
Cliches aside, I think you are right about the poor decisionmaking: IBM decided to chase the biggest, most bureaucratic market, healthcare, rather than pick a market that would actually be somewhat receptive to black-box services.
Watson (the Jeopardy playing machine) really was an amazing demo for it's time (2010) and before the rise of neural networks in NLP it was one of the best demonstrations of a semi-practical system.
The "This is Watson" special edition journal[1] taught me more about traditional NLP pipelines than just about anything I've read before or since.
It was an amazing demo, as far as marketing was concerned. Now whether the technology behind it was as revolutionary as they made it sound, that's probably more controversial.
It was certainly a stye in the eye for Google (“organize the world's information and make it universally accessible and useful”), whether facbricated by marketing/advertising or not.
But it wasn't really putting itself out there as a Google competitor. IBM wasn't saying, "buy this to make a better search engine", so I never once drew the Watson vs Google comparison when it all went down in 2011. Watson seemed to want to market itself as tech to drive expert systems instead (though it never really panned out) and now 9 years later, Google Assistant can answer questions of astou ding complexity, so whatever research Google was embarking on in 2011 has since born much more fruit anyway.
That's somewhat narrowly defined considering they're really the only remaining producer of mainframes. I'm not sure if that's literally true but it may as well be.
There are some things that mainframes are legitimately better at, and some things that POWER is better at than x86, and so on. AMD and Intel benefit from incredible economies of scale that IBM could never dream of, and that's mostly where IBM loses -- the price is a lot higher and the volume a lot lower on the IBM side.
I wonder if we'll see IBM get into video games before too long. Or buying a lesser known streaming company like Mixer? It would seem like an odd fit, but it wouldn't be too surprising.
...and it's been integrated into a couple of MS games, too. Forza Horizons 4 has Mixer integrations, and even gives you a bonus experience multiplier for streaming or something like that.
edit: sorry. I misunderstood you. Some projects are delivered with subcontractors, of course. But I can tell you from inside that we still have a healthy GBS, GTS and Cloud Services, delivering projects.
Go to https://en.wikipedia.org/wiki/List_of_mergers_and_acquisitio... and sort the table there by value, descending. "IBM" is a shorthand for more (un-dissolved!) subsidiaries at this point than GE or Samsung.
IBM itself is just a consulting "solution provider" company that provides "solutions" in terms of the products and services of its subsidiaries. They're never going to look like they're on the cutting edge of some space, unless that's what one of their (equally-stodgy) enterprise clients wants to pay them to do. (And even then, why build when you can buy? IMHO IBM would only ever build a new technology at this point if there was no existing company out there in the market, with that tech IP, that they could just acquire.)
Certainly, there's the Watson AI lab, though I don't think anyone's expecting much of it at this point; nor, importantly, was anyone inside IBM ever expecting much of it. Watson (the lab specifically, not the brand strategy of calling all their cloud OLAP SaaS products "Watson") is more a nod to their clients to say "of course we're looking into AI", and a pool of people competent enough that they could clone the existing ML approach of one of their competitors if one of their clients demanded it.
IBM could be a really innovative company if they rearranged their talent pool, certainly. But innovation—getting ahead of the market and so needing to educate the market on the problem their products solve—hasn't been IBM's business model since the 1950s. IBM makes money by listening to clients' "unique" needs, and then meeting them with "custom" solutions at low CapEx cost, by using an ever-larger flywheel of existing solutions pre-tuned to almost fit every possible business use-case.