The massive rounds of layoffs in the US over the past 3 years mean there are a ridiculous number of software engineers looking for jobs. AI has compounded this by automating applications such that a single opening might get multiple thousands of submissions.
Your best bet to find a job in this market is to have some connection to the hiring manager. It might be a friend of a friend of a former colleague. Or both having membership in the same semi-open community. But you need a way to say “hey, I’m a real human being and especially interested in this job; please at least take a look at my resume!”
The other thing I had to realize before I could get an offer in this market is that there are so many overqualified candidates it’s almost impossible to get a stretch role and very difficult to get one at the same level you were last working at. If you just need an income for now, consider looking for something you’re significantly overqualified for but that gives you exposure to something new in a different dimension (new industry, for instance). Then once you’ve got that job and have restored your savings, you can look for something with more growth while still employed. Maybe some day the market will even bounce back, tho I wouldn’t count on it.
If possible, go to local meetups for whatever type of role you are in/interested in. The current environment while very different from the 2000’s dot com bubble bust, has certain similarities, and at that time, the only way to really find work was through relationships. I know at that time I ended up switching from being a software engineer to desktop support for about 6 months just to stay employed, especially since it was the only job available in my friend group.
I think that the author has a point that LLMs need to make sense (and money) in the context of products.
Google’s Gemini integration from a consumer standpoint seems to be doing it right, even though Gemini on the developer side could mean many things.
We’ve probably been on the cusp of the transition from the exponential exuberance on LLM hype towards building products that make money. Still going up, but the hype is slowing.
The density required for solar is also much lower - the coordination between different land parcels and routing power and getting easements increases the time required vs. on prem gas turbines.
Gas plants are not bad… but imagine 400 MW of gas plants in a concentrated area. You’ll always have NOx and SOx by products whenever you’re burning gas.
Direct air capture imo can’t escape the scaling problem - when the feedstock has CO2 at ~400 ppm the economics simply won’t work out despite various oil companies backing one off systems around the globe.
Capturing CO2 at the source (power plant, etc) would be simpler to reach economic viability but without incentives it’s dead on arrival. I believe the IRA infra bill had put a price ~$50/ton of CO2 captured.
But we still need to remove all the excess co2 that we released into the atmosphere since the start of the industrial revolution if we want to reduce the temperature back to what it was before we started disrupting the natural state of the plane.
We and previous generations took out a loan and the payment is coming due.
Because of the framing about degrees in celcius change people are thinking in small numbers, like "oh, it's just 1.5'C over normal. oops, we missed that, well maybe we'll get it at 2.0'C. They don't realize that if we want normal we ahve to reduce the temperaure and to do that we need to take that c02 blanket off that we've been tightly wrapping around our collective bodies for decades.
And that endeavor is nearly unfathomable. Think of all the energy used by humanity since the industrial revolution and the energy we're going to be producing in the time period that we attempt to sequester the previously poduced C02. All of that needs to be accounted for.
And then there's the surplus energy roiling around in the system now, and the collapse of food webs.
I don't see how we get our way out of this in the next 50 years.
With ice caps melted off, just removing all the excess CO2 isn't even enough since with that reflective surface gone, more energy from sunlight stays in the atmosphere than previously when more of it was reflected back into space instead of nowadays being absorbed by the ocean.
Absolutely. People seem to think that we just need to recycle more, seal some cracks in the house with foam, install some solar panels, and buy an electric car.
They underestimate the scale of the intervention that will be required to stave off the potential end of human civilization as we know it. If we have any hope of continuing to live at something resembling the quality of life that we've grown up in it will require radical science fiction like developments.
We're going to need things like space based solar shades to regrow glaciers and icepack, advanced breeding and cloning and ecosystem engineering to reconstruct collapsing food webs, and I think the big picture thing is that we're going to need to engineer people to reduce susceptibility to addictive food and manipulative marketing.
That’s true. It’s more of a policy issue that’s like carbon credits… nice on paper but a big nothing burger. Look at F1 and Porsche talking about sustainable synthetic fuels.
When you compare round trip efficiencies and economics it makes sense to just not burn the hydrocarbons to begin with.
If stored near a populated area, hundreds of thousands could be kill, including all animals and insects, in a matter of minutes if the "vault" has a catastrophic failure. I would rather live near a nuclear waste site than a CO2 Site.
Well as far as storing it goes, if you can capture it, turn it into a solid and stick it in the ground.
Imagine you were growing a huge biomass that you harvest, dry out, and then store. We know how the bacteria and processes that stripped co2 from the atmosphere in the past, we just need to do that in a big way. Good thing we have places on earth that are huge and flat and growing algae won't be a problem.
And then we complement that with green energy and an attempt at net zero.
Chemical reduction of CO2 is a very hard problem since it takes a lot of energy. There's an enormous amount of research in this area. Storing it costs money, so most of the research focuses on turning the C into a useful product.
A failure in containment at a nuclear waste storage site means danger for people on the site itself but it is easily detected by monitoring equipment and can be repaired, the waste is solid and can't spread easily. A failure in containment of a massive quantity of pressurized CO2 would be significantly more dangerous and probably a lot more likely to happen given the frequency of accidents
Capturing CO2 at the source will always be worse than removing the source. At the same time, capturing CO2 from the air will stay necessary until we do it.
DAC is unsustainable, and shouldn't be a primary offset mechanism. What would be more sustainable is less net emissions and using something else like oceanic biomass CCS by taking advantage of blooming kelp and/or phytoplankton and then processing it for sequestration using floating, automated harvesting rigs.
The re-edited title frames this as an anti-drone system but this was foremost developed as an anti-rocket system.
Hamas and Hezbollah MO since the 1990s was based on bombing Israeli towns with statistical rockets and this system is supposed to reverse the cost equation (cheaper than those cheap rockets)
The rockets are very imprecise, but a large number of them, hitting the territory of a town, will deal damage, bodily harm, and death at random, due to statistics. It's Monte Carlo bombing of sorts :(
They say it's first operational system in it's class, but it seems very similar to the Australian Apollo system, with Apollo being able to go up to 150kW
It's also similar to the British DragonFire and US HELIOS
I think the major difference here is that the Iron Beam is operational, as in finished trials, delivered to an armed force and actually was in active use in the previous war for more than a year
Good question, probably depends a lot on how much energy actually makes it to the target some distance away. And then how much is actually absorbed. Probably depends more on the power density then, rather than total power?
Can't imagine they get a very small spot at multiple km unless they use gigantic lenses or multiple independent laser focused on the same spot
I also wonder the extent to which the effectiveness is reduced by painting the projectile white or wrapping it in aluminum foil. Maybe 100kw is so large that it simply does not matter at that power level.
It'll get a lot of time to react at that energy as it's not going to "instantly" fry anything*. That's probably less energy/m2 than consumer heat guns, especially if consider that these drones are likely going to get sprayed in reflective paint. Easy defense for the drone would be just: get into a spin to get roasted evenly -> shut off -> fall for a few hundred meters, cooling using air that rushes by to counteract the laser further -> catch itself once it lost the laser.
That would force these laser systems to point each drone until it either visibly goes up in flames or impacts the ground (which means you also need to be able to track them all the way down), otherwise you can't be sure it won't just snap back to life once you started engaging the next drone.
I don't feel like 10kw/m2 would be anywhere near useful. It's gotta be more than that.
* Stadium floodlights aren't going to instantly grill any bird that flies in front of them either, and they reach that ballpark.
There is 400 amp residential service you can get 80 amp 19.2 kw level 2 chargers.
You would need 5 80 amp charger to approach 100kw but with other loads in a large house, I have seen large HVAC systems and elaborate pools with lazy rivers etc that can add up very quickly which is why they had 400 amp service.
100kw isn't really that much, a modern EV can put out 3 times that from its battery pack into the motor for short bursts and easily sustain 100kw until drained.
480v 200 amp 3 phase commercial supply can provide 100kw continuous and would be some thing used in a medium sized office building.
As the sibling comment notes, these days 400a residential service is available as an option in many places.
One home actually consuming close to 400a is pretty rare, but it's possible mainly in gas-free builds, if using things like electric tankless water heaters (a bit niche) in addition to multiple EV chargers, a range, dryer, etc.
Maybe a better way to convey that 100kW is “small” is to point out that industrial sites all around us, such as smaller datacenters, are well into the MW range.
Huh, to what degree is this technology gatekept by battery advances?
A few decades ago lasers were dismissed because they involved chemical reagents for high power and explosive capacitors for even low-power applications.
> Huh, to what degree is this technology gatekept by battery advances?
Not too much. The power delivery was doable even 15 years ago. It would have just been more expensive and heavier.
The bigger issue I believe would have been the lens and tracking capabilities. For the tracking to work you need some pretty good cameras, pretty fast computers, and pretty good object recognition. We are talking about using high speed cameras and doing object detection each frame
> The power delivery was doable even 15 years ago.
Not really. It took a long time for solid state lasers to make it to 100KW. That's the power level military people have wanted for two decades.
Megawatt chemical lasers are possible, and have been built. But the ground based one was three semitrailers, and the airborne one needed a 747. Plus you ran out of chemicals fairly fast.
I took 'power delivery' to mean the systems that facilitate driving the energy into the weapon, not the beam itself -- although now under consideration of the technology I think we should probably avoid the use of the phrase 'power delivery', without a projectile being involved that's essentially the entire concept.
A 100KW generator is no big deal. It's a truck Diesel engine coupled to a generator. Trailer-mounted, it can be towed with a pickup truck. It's a standard rental item for larger construction projects.
A 100KW laser is a big deal.
The big problem with this as an anti-drone weapon is that, unlike artillery shells or unguided missiles, drones can operate close to the ground, and the laser needs line of sight.
If these things are even 50% efficient, then power delivery is really not a problem these days. Most EVs have no problem delivering 200kW for quite a few seconds at a time, limited mostly by components getting warm. Higher-end EVs are generally rated for 300-500kW.
It would by amusing to see one of these lasers mounted on an EV, possibly with a small range extender to recharge it on the go.
It’s missing almost all technical details, which seems fishy to me. But I’m sure this defense company is honest and has a system that works great and so that’s why no technical details are needed. /s
When you're playing with nukes it actually is rather effective, not from a standpoint of chaff (you don't bring it) but the ionization of the nuke makes a radar blocked zone and the following missile is going very, very fast--makes a bunch of progress while the defenders are blind. It's also why we don't like nuclear anti-sub weapons--the dead zone lasts for hours, there's no way to know if you actually got the target.
But a drone is small and slow. You'll need an awful lot of drones to punch through defenses this way and the whole thing goes out the window when the laser pops drones farther back in line. And chaff only denies a small area and for a short time.
You’re right for ambush drones of the sort e.g. Hamas could launch. For the ones that would stream in from Iran, which Israel needed American help defending from last time, I’m not sure that’s the case.
I mostly work with arrays in numpy but sometimes I get stuck with a problem that needs a for loop - only two things I can think of are parallelize the for loop or use numba/jax JIT functions and GPU acceleration.
Both don’t just work out of the box like Julia or MATLAB’s “parfor” loop, but seem to work well enough for non trivial for loop cases.
Sure. To be fair, having gone down the path of porting and testing a problem to numba, it might be easier to just jump to Julia if you want to focus on the problem more than the implementation.
More so that the 1% have fewer loopholes, not the other way around.
Regular folks filing taxes don’t have particularly difficult returns, but if the government already knows what you owe/get back, why waste everyone’s time? Obviously deductions exist for a reason, but standard deduction folks shouldn’t have to file to confirm that their number matches what the IRS already calculated.
At least, that way you know it’s not the remote work portion that’s keeping you from a job.
I’m in the US and everyone I’ve talked to who wants to move have been discussing the challenges of getting a foot in the door anywhere.
reply