4 ms·

DARPA has a rich history of doing fundamental research but today its largely composed of engineering tasks which for whatever reason arent built by the warfare

DARPA has a rich history of doing fundamental research but today its largely composed of engineering tasks which for whatever reason arent built by the warfare centers. The days of doing fundamental research are mostly gone in the organization. Name something in the AI/ML space that came out of DARPA in the last 15 years. CNNs- no, deep learning - no, attention - no, LLMs - no, test time training - no. Its a good engineering org these days building cool things but there's nothing special about having darpa build it other than they provide funds. The research focused has significantly waned over the last 20 years which is a shame as there are LOTS of fundamental AI/ML problems which no one is studying in industry.

whiplash451 2y ago
AI/ML is a solution, not a problem. DARPA’s mission is to define the problem space, not the solution space.
- ipunchghosts 2y ago
  YES! I should add that they should be thinking about intelligence generally, not AI/ML.
- jacobr1 2y ago
  A good example of this was the grand challenge with autonomous vehicles. It spurred all sorts of AI and Robotics related innovation that spun into different directions.
  - whiplash451 2y ago
    This almost made it to my previous reply, but the DARPA Challenge was 18 years ago. There might be other ground-breaking challenges that DARPA did since then, I'm just not aware of them.
    - ipunchghosts 2y ago
      I believe the contribution here is overstated. None of the tech used for today's AVs relies on anything developed for grand challenge, including any principles they derived. The leading tech today was derived purely from industry outside the dod space and cifar.
      - whiplash451 2y ago
        That's not the point. The point is to put it front and center for the tech companies and the general community (think of it as a call for proposal). Would Google have invested into self-driving if DGC had not happened? Very unclear. Note that the initial Waymo team was a direct draft of the winning Stanford team at the DARPA Urban Challenge'07.
        ipunchghosts 2y ago
        I think googles could care less about what darpa is doing. How would I know? I'm a darpa PM and google isn't reaching out to me or my team. I think what you are seeing is correlation but not causation. Stanford was working on self driving long before darpa, same with cmu. The darpa money was just a bonus but I don't think it pushed anything ahead. After the first few years of the darpa challenge, many teams were able to finish. Yet here we are in 2025 and I cam get bounding box detections on video to not jitter or drop between frames because of some unlucky pixel combination. Humans have zero issues and aren't even able to perceive these minute differences. My guess is vision will be solve in about 18 months using methods developed purely in industry.
      - kelipso 2y ago
        You need an entire community of researchers (both academic and non-academics) working on a field for industry to produce tech. It was definitely not a pure industry thing, almost no tech is.
        ipunchghosts 2y ago
        Sure, but none of that was developed by dod in any real way. A litmus test to consider is that the dod still halls belig ml problems but you don't see phds funded by dod going to the dod. The good ones all go to industry because they know thier reseaech stops as soon as they start working for a contractor or warfare center. The money excuse is just a patsy. The government has plenty of money to pay these ppl if they can produce which they do well in industry. It's simply not a focus of darpa to do basic research in learning and intelligence in a meaningful way.
_yb2s 2y ago
> there's nothing special about having darpa build it other than they provide funds DARPA chooses real world engineering/technical problems, and then works closely with the external grantees that develop possible solutions, to solve the problem together - rapidly. Fundamental discoveries often come out of that kind of focused well resourced problem solving, but it's not really the goal.
- ipunchghosts 2y ago
  That's what ousd is for.
mometsi 2y ago
Edit: Whoops. I incorrectly believed that Caffe was developed with DARPA support. I see DARPA acknowledgments on papers written at BLVC by some of the same authors, but not on Caffe itself.

whiplash451 2y ago

AI/ML is a solution, not a problem. DARPA’s mission is to define the problem space, not the solution space.

ipunchghosts 2y ago
YES! I should add that they should be thinking about intelligence generally, not AI/ML.
jacobr1 2y ago
A good example of this was the grand challenge with autonomous vehicles. It spurred all sorts of AI and Robotics related innovation that spun into different directions.
- whiplash451 2y ago
  This almost made it to my previous reply, but the DARPA Challenge was 18 years ago. There might be other ground-breaking challenges that DARPA did since then, I'm just not aware of them.
  - ipunchghosts 2y ago
    I believe the contribution here is overstated. None of the tech used for today's AVs relies on anything developed for grand challenge, including any principles they derived. The leading tech today was derived purely from industry outside the dod space and cifar.
    - whiplash451 2y ago
      That's not the point. The point is to put it front and center for the tech companies and the general community (think of it as a call for proposal). Would Google have invested into self-driving if DGC had not happened? Very unclear. Note that the initial Waymo team was a direct draft of the winning Stanford team at the DARPA Urban Challenge'07.
      - ipunchghosts 2y ago
        I think googles could care less about what darpa is doing. How would I know? I'm a darpa PM and google isn't reaching out to me or my team. I think what you are seeing is correlation but not causation. Stanford was working on self driving long before darpa, same with cmu. The darpa money was just a bonus but I don't think it pushed anything ahead. After the first few years of the darpa challenge, many teams were able to finish. Yet here we are in 2025 and I cam get bounding box detections on video to not jitter or drop between frames because of some unlucky pixel combination. Humans have zero issues and aren't even able to perceive these minute differences. My guess is vision will be solve in about 18 months using methods developed purely in industry.
    - kelipso 2y ago
      You need an entire community of researchers (both academic and non-academics) working on a field for industry to produce tech. It was definitely not a pure industry thing, almost no tech is.
      - ipunchghosts 2y ago
        Sure, but none of that was developed by dod in any real way. A litmus test to consider is that the dod still halls belig ml problems but you don't see phds funded by dod going to the dod. The good ones all go to industry because they know thier reseaech stops as soon as they start working for a contractor or warfare center. The money excuse is just a patsy. The government has plenty of money to pay these ppl if they can produce which they do well in industry. It's simply not a focus of darpa to do basic research in learning and intelligence in a meaningful way.

_yb2s 2y ago

> there's nothing special about having darpa build it other than they provide funds DARPA chooses real world engineering/technical problems, and then works closely with the external grantees that develop possible solutions, to solve the problem together - rapidly. Fundamental discoveries often come out of that kind of focused well resourced problem solving, but it's not really the goal.

ipunchghosts 2y ago
That's what ousd is for.

mometsi 2y ago

Edit: Whoops. I incorrectly believed that Caffe was developed with DARPA support. I see DARPA acknowledgments on papers written at BLVC by some of the same authors, but not on Caffe itself.