Pentagon's New AI Chief Vows to Crack 'Bureaucratic Inertia' on Tech Advances (bloomberg.com) 24
The Pentagon's new head of artificial intelligence wants to speed up technological modernization after an onslaught of what he calls "valid" criticism from recently departed senior leaders who expressed frustration at slow progress. From a report: Craig Martell, who was previously head of machine learning at Lyft and Dropbox and led AI initiatives at LinkedIn, told Bloomberg News in his first interview since starting his job as the Pentagon's chief digital and artificial intelligence officer that he wanted to make progress despite the department's labyrinthine "bureaucratic inertia."
Martell's arrival is a boost for the Pentagon, which is seeking to attract expert talent from the private sector. Martell, who said his first day at the Pentagon on Monday was "overwhelming," added he had taken a "not trivial" pay cut to do the job. "It's not my goal to come in here and change the entire culture of the DoD. It's my goal to demonstrate that with the right cultural changes, we can have really big impact," he said, adding that the opinions of senior leaders who had left were "mostly correct."
Martell's arrival is a boost for the Pentagon, which is seeking to attract expert talent from the private sector. Martell, who said his first day at the Pentagon on Monday was "overwhelming," added he had taken a "not trivial" pay cut to do the job. "It's not my goal to come in here and change the entire culture of the DoD. It's my goal to demonstrate that with the right cultural changes, we can have really big impact," he said, adding that the opinions of senior leaders who had left were "mostly correct."
Vows to Crack 'Bureaucratic Inertia' (Score:2)
The Persian kings tried it, the Egyptians, the Greeks, the Romans, they all failed.
Re: (Score:1)
Re: (Score:1)
Indeed! They probably should change their goal to modularize so that they can purchase private sector technology and hook it in where needed rather than try to do it all in-house.
Re: (Score:1)
Did you mean this sarcastically? Basically the problem is they just buy everything from Microsoft and don't even bother with due-diligence, so they're about as hardened to international hacking threats as your average uneducated senior citizen.
Two ways to avoid slow bureaucracy (Score:2)
There are two ways to avoid the slow-motion inertia of a giant bureaucracy. Both work.
One is to not have a giant bureaucracy, but rather many small teams. If you call OSHA about a workplace hazard they may come check it out in about 6-9 months. After the paperwork winds it's way through many offices in many states. If you call the local fire department because that workplace hazard is now a workplace fire, they'll be there in 6 minutes. They don't have a giant bureaucracy; they have a five-person team at t
Re: (Score:3)
"The other option for efficiency, to avoid bureaucracy, is dictatorship. Kim Jong-un says "shoot that guy", they shoot him, right then and there. No waiting years for appeals at various levels. Kim speaks, it's done, the guy is dead right away. Kim Jong-un says "send $4 million to XYZ Corp to buy missile bodies", it's sent right away. There's no process to make sure XYZ corp isn't owned by Kim's cousin. No months spent on competitive bids. There's no acquisition process spending months to ensure the owners
Doesn't make sense, print money. (Score:1, Troll)
Doesn't make sense that the government who can essentially print money for use by spy/defense with secret black budgets can't just fricken pay for top talent.
My god. Throw some 300-400K+ offers out there and people may actually jump ship.
This is another spin on school, where the people there aren't so much the examples of the greatest but rather those who couldn't make it in the private sector leaping into a comfy public gig.
Real talent is often held by people who expect above-average compensation for use o
Re: (Score:2)
There's a lot of people who simply won't work for the gov't. No stock options, no bonuses. 535 ignorant, if not downright stupid, managers outside of your own management chain. Sure, the benefits are good, but a lot of people want more autonomy in their work, which you don't get in a gov't job.
Re:Doesn't make sense, print money. (Score:5, Insightful)
Considering the failures of a multitude of companies who are paying through the nose for supposed A-grade people, your comment doesn't hold water. One quick example, that actively exploited zero day issue Microsoft still hasn't corrected [slashdot.org] in nearly two months. And yet, their programmers are some of the highest paid in the industury and recently the company said it's increasing salaries again. I don't think they're getting their money's worth based on not only this flaw which hasn't been corrected, but the weekly notices of other such issues.
If you want to go a bit further back, look at the revolving door at HP and how those hacks got paid oodles of money yet still drove the company into the ground. Yahoo! ring a bell? Let's not even go into Wall Street firms who repeatedly come hat in hand to the U.S. taxpayer to bail them out from their incompetence. How much do you think hacks like Jamie Dimon and Lloyd Blankfein got paid for those disasters?
I could go on, but this nonsense that people need to be paid exorbitant amounts of money and perks to produce when the overwhelming evidence shows there is no relation to performance is in the same realm as trickle down economics. Neither have worked but we keep repeating the same mistakes. It's like deja vu all over again.
Re: (Score:2)
You will only get wannabes and B-grade people without ponying up the $$$$.
Considering the failures of a multitude of companies who are paying through the nose for supposed A-grade people, your comment doesn't hold water. One quick example, that actively exploited zero day issue Microsoft still hasn't corrected [slashdot.org] in nearly two months. And yet, their programmers are some of the highest paid in the industury and recently the company said it's increasing salaries again. I don't think they're getting their money's worth based on not only this flaw which hasn't been corrected, but the weekly notices of other such issues.
If you want to go a bit further back, look at the revolving door at HP and how those hacks got paid oodles of money yet still drove the company into the ground. Yahoo! ring a bell? Let's not even go into Wall Street firms who repeatedly come hat in hand to the U.S. taxpayer to bail them out from their incompetence. How much do you think hacks like Jamie Dimon and Lloyd Blankfein got paid for those disasters?
I could go on, but this nonsense that people need to be paid exorbitant amounts of money and perks to produce when the overwhelming evidence shows there is no relation to performance is in the same realm as trickle down economics. Neither have worked but we keep repeating the same mistakes. It's like deja vu all over again.
I think some of this is a misunderstanding by companies about how spending money, vs investing in the company, actually works. You can have the best coders but crap managers and produce crap projects. You can throw tons of money out there and hire useless people if your HR doesn't know what makes a good employee. You may have the best of everything, but if you don't keep up with training people on emerging technologies and moving the company forward, the company will stagnate and become meaningless. Yes
They get what they want (Score:3)
> One quick example, that actively exploited zero day issue Microsoft still hasn't corrected in nearly two months.
You're assuming that fixing these types of things properly is Microsoft's number one priority. It's not. Microsoft's people are doing what management wants them to do.
Microsoft's business and financials is divided into three sections, which you can see on their annual report:
Azure
Office365
Gaming and Other
This isn't the Bill Gates years, this isn't Bill Gates Microsoft. This is Nadella's Micro
Re: (Score:2)
This isn't the Bill Gates years, this isn't Bill Gates Microsoft. This is Nadella's Microsoft. You're talking about Windows, which is now part of "other". Nadella is spending a lot of money on top tier people to build what Microsoft is focused on - Azure and Office365.
I'm not sure how "this isn't gates' microsoft" is supposed to be relevant here. Microsoft never prioritized fixes of any kind. Windows is still enormously relevant to their strategy, though. It's still a vendor lock-in issue. They make it easy to use Azure with Windows tools so the two support one another.
Re: (Score:2)
Under Gates, the entire company was there to support Windows.
For anything the company did, the question was always "how does that help Windows?"
Not so with Nadella. Under Nadella, Windows is literally lumped in the "gaming and other" category. It's "other".
Re: (Score:2)
> One quick example, that actively exploited zero day issue Microsoft still hasn't corrected in nearly two months.
You're assuming that fixing these types of things properly is Microsoft's number one priority. It's not. Microsoft's people are doing what management wants them to do.
Microsoft's business and financials is divided into three sections, which you can see on their annual report:
Azure Office365 Gaming and Other
This isn't the Bill Gates years, this isn't Bill Gates Microsoft. This is Nadella's Microsoft. You're talking about Windows, which is now part of "other". Nadella is spending a lot of money on top tier people to build what Microsoft is focused on - Azure and Office365.
The work you are talking about is done by the Quick Fix Engineering team. That's their actual title, QFE or quick fix engineering. Aka bandaid, duct tape. They bandaid that vulnerability every month.
Microsoft doesn't need to fix that exploit. Because the loyal users won't do anything - they will continue to use Microsoft, and nothing will pursuade them otherwise.
So I'm not saying you are wrong about the concept that money is the sole arbiter of quality, using Microsoft as an example doesn't work.
Re:Doesn't make sense, print money. (Score:5, Interesting)
We were working on a program that involved manned systems, and there was a safety system (fire suppression) on board. The system was designed in 1942 (this was 2006). The system was custom built because no one made it anymore, and commercial equivalents had moved on in terms of whole generations of technology. We pushed for moving to a newer system, proven in the commercial world, but it took us 18 months and millions of dollars to prove it thoroughly. Why? The Warrant Officer. There are a group of minor officers who's sole role is to be responsible for certain systems. In this guy's situation, he was responsible for this fire safety system and thus the lives of military personnel. This system worked; a new one was a risk and might kill a soldier. Why change it? He had no incentive to upgrade or save money, but he had every incentive to save lives. That was the structure.
In another situation the DoD wanted to extend a contract for a given system; the contract started in 2000 and it was 2008. Their mandate was all new systems were built identical to the original systems bought. All of that makes sense, except for software. In 2000, Windows XP was the standard; in 2008 it was being phased out. Actually, did you know that the majority of US Navy ship systems run Windows XP still? Because change is disruptive.
Because of their slow pace of change, the DoD then requires things that most commercial entities would never give up or do. Like complete design schematics, or the source code, or manuals for highly technical systems written to the reading level of a C- High school graduate and not a trained engineer.
The unfortunate thing about the DoD is it's dealing with a series of issues that are intrinsic to its very existence:
1) It has tens of thousands of new recruits every year. It's constantly training people on new systems or old systems, and many of them are not super well educated. Consistency is critical to mission performance, so if a technician is coming in to a new theater and the equipment is totally different than what he trained on, it affects mission performance. They've never found a good way to keep up consistent training with consistent technical change; the tech/software world in particular moves exceedingly quickly compared to the pace of training the military must deal with.
2) The budget/acquisitions cycle - the DoD spends hundreds of billions of dollars per year, on a per-year budget driven by political wrangling and nonsense, often funded via continuation amendments rather than passing a formal budget. Most of their programs require multi-year expenditures, but they cannot plan for that. Sometimes they're forced into buying stuff they don't need because of some Senator's pet project to drive jobs to his state, or the cancellation of planned orders for equipment because some Congressperson got it in their heads to fund some social program and take the money from the DoD budget, which of course drives up overhead and thus unit cost. Even when they do something reasonable, like JEDI, government projects are open to critique from the bidders, resulting in a lawsuit holding back the project. By the time the lawsuit is settled, the specs are obsolete and the program means nothing.
3) This part is changing, but changing slowly. The military is missing a conceptualization of what tech means to it. In most top brass' view, software is something that helps hardware operate. They think in terms of men and material, ie soldiers and tanks/ships/planes. Software makes it easier for soldiers to drive tanks/ships/planes. But we all know, and to some degree the military knows, that software is now something devoid of the physical hardware it's attached to. They don't yet know what this means, and that lack of understanding drives their acquisition programs. their
Re: Doesn't make sense, print money. (Score:2)
tech bro redux (Score:1)
Not as long as their goals are stuck in the past (Score:2)
The military can only support the MIC and preserve the status quo. It can't bring us anything new. As long as our government is strongly authoritarian towards citizens while being in the pocket of corporations it will always retard technological process, and it is that way and it is getting more strongly that way. The big guys always have incentive to slow down improvements because they might come from someone else.
AI-GOvT (Score:2)
Introducing AI-Governmental Overt Technology.
Why not crack bureaucratic inertia on tech advances with ... tech advances? Any large system is complex and difficult to manage en masse. But AI is perfect for that. We need to replace Congress with a network of AIs. Faster, cheaper and more accurate - when before you could only have two out of those three.
Re: (Score:2)
Any large system is complex and difficult to manage en masse. But AI is perfect for that.
Yes, a computerized solution makes it easy to link in a payment system.
Here comes Skynet (Score:2)
So he wants to nuke the place and start over? (Score:2)
Sounds like a good idea to me...
Military pays decent... (Score:3)