cal5k
We're building the machine that builds the ships - a high-mix robotics platform for autonomous shipbuilding.
If you love making giant industrial robots do things they were never designed to do, we'd love to talk. We built our own software stack from scratch, from motion planning to physics-based simulation, so you'll get to work on really interesting problems that are usually a black box in an industrial setting.
The role: https://valstad.com/careers/senior-robotics-software-enginee...
Other positions: https://valstad.com/careers/
Most of the crime guns used in Canada are actually illegal to purchase in the country. They're trafficked by organized crime because there's a lucrative market for them - see note about failure to lock up violent repeat offenders. Massive court backlogs and "bail reform" mean criminals pending trial for serious offenses are often back on the streets in days.
> You Imagine you alone with a AR-15 vs 5 terrorists with AK's and you win, right?
No, I'm imagining deterring specific, targeted kidnapping attempts by armed kidnappers.
Even if the police magically are able to intervene in time you're not getting former JTF2 members like PPS. You're getting people who probably hit the range twice a year and have likely never drawn their firearm on the job before. They may or may not even have a patrol carbine (C7/C8) in their squad car. That's no shade on Toronto Police, that's just the reality of their jobs. If ERT (SWAT) gets called out it's only happening after you've been kidnapped and moved to a second location - assuming they can locate you in the first place.
Many legal/licensed conceal carriers I know in the US train more often than the average police officer. In that scenario the mere presence of a firearm - in addition to other security measures & situational awareness - often deters would-be attackers without a shot being fired. This is the ideal scenario since no reasonable person should want to get into a gunfight.
> your imaginary hero self-defense stories.
This isn't imaginary, this literally just happened as per the article to someone I know personally in the same industry.
Cooperating with kidnappers also does not guarantee that you will come out alive. Targeted killings are a thing that happens:
https://www.cnn.com/2022/12/17/us/toronto-billionaire-sherma...
Two things are simultaneously true in Canada right now: 1) the government has become unwilling/unable to actually keep repeat violent offenders in jail (and has totally failed to prevent them from acquiring illegal firearms), and 2) the law essentially outlaws the use of any tools (firearms, pepper spray, knives) for the express purpose of self-defense.
It's functionally impossible to get a concealed carry permit in Canada. They just don't issue them, even if you have a restricted firearms license, even if you're a high-profile individual, and even if there's an immediate threat to your life. You have to rely on police and, if you can afford them, expensive private security who probably can't carry a firearm either. I should note that the kidnappers in this case (plural) were all armed.
The politicians, of course, are protected by the armed Parliamentary Protective Service. They just don't believe you deserve similar protection.
How about in a Tesla?
They are approaching this from two fundamentally different directions.
It's bizarre how people are so adamant about this when they haven't even tried it. Every passenger I've shown this to have had their minds blown.
From the article. Have the people saying this actually tried the most recent FSD build?
I used it yesterday, and it reacted flawlessly to road construction that blocked a lane, then had cones demarcating a lane that didn't follow existing road markings. Then it saw that a driver wanted to turn left from parking lot while approaching a stop light and made space for the driver to turn out. It also flawlessly executed a merge involving turning into a suicide lane and waiting for traffic to pass.
It's very good at adapting to changes and can work on roads its never seen before. There is no other production car that can do what FSD can do currently.
Does this mean it's ready for full autonomy tomorrow? No, but I could use supervised FSD today if I wanted to drive 1600 miles from Austin to Toronto with hands mostly off the wheel.
The pace at which FSD is now improving is quite remarkable. In the long run it's clear that this approach is going to be WAY more flexible than Waymo's, because Tesla has so many more vehicles on the road collecting so much more data, and throwing vast amounts of compute at vast amounts of data is basically the best way to see rapid improvement for the foreseeable future.
But sure, "concepts of a plan".
If you're a large enough entity operating in multiple jurisdictions around the world, you're exposed to such a byzantine labyrinth of laws and regulations that it's impossible not to be in violation of some rule or law, somewhere, at all times.
Regulators know this. They use discretion to decide when and how to enforce these rules. Some are seldom enforced, others enforced with reckless abandon depending on the administrator, market conditions, political winds, etc.
Every regulated financial institution, for example, and I mean every single one, will be fined for something if it is big enough and has been around long enough.
Basing fines on worldwide revenue is a ridiculous way to ensure absolutely no risk-taking happens, because innovation happens at the margin. Tomorrow's massive industry is today's grey area.
Vitalik strikes me as being very naive about the process of policy-making. All politicians are, by definition, power-seeking individuals. They're seeking the power afforded to them by a seat on city council, as governor, a member of congress, etc. This is simply the nature of the political process. They may even be seeking power in pursuit of a noble objective, but they are seeking power nonetheless.
Realpolitik as a non-politician means aligning yourself with whoever best represents your interests at that particular point in time. One election cycle that may be Party A, but in the future it could be Party B. It could differ at the federal, state, or local level. Frequent elections and term limits ensure that turnover will happen, so change will be the only constant. Flexibility is required.
Waiting for the perfectly ideologically-aligned candidate sounds nice, but in reality it means you'll be waiting forever. All politicians are flawed, and many will simply tell you what they think you want to hear. Their behaviour is amenable to incentives, however, which is why donations are such an important - if maligned - mechanism for accountability.
Marc & Ben are being practical. They've observed that the behaviour of one party has run directly counter to their interests over the last several years and they're aligning themselves with the party that is at least proposing a better set of policies.
That's just... how this stuff works and has always worked.
Why are regulators still attacking stablecoins, for example, when they represent one type of innovation that could actually lower transaction costs? Creating a legislative framework that encourages innovation rather than stifles it would make a lot more sense than trying to micromanage fees.
This is a nakedly partisan substack account. Does this really meet HN's quality standards?
Oh, and as a nice little bonus they also curb substance use disorder!
As for it being a feature, you're right - I think the vast majority of legislating should be done at the state and local level. This isn't a secret, it was covered in the Federalist Papers in the 18th century.
What's your argument in favor of federalizing all aspects of law in a large and heterogenous country? Why does the federal government need to force Texas to ban non-competes when Texas has decided not to do so but California has?
What if the next government decides to force California to un-ban non-competes with a new rule issued by Executive Order? Do you not see why this is an unworkable and brittle approach?
That's not what happens, though. Agency heads are political appointees who take their marching orders from the executive (president). They engage in poor-quality rule-making all the time when it's politically advantageous for the president and/or his party.
Coming from a country with a Westminster system where federal legislation is relatively easy to pass, I strongly believe it's a feature of the US system of government that it's a herculean task for congress to pass new laws and that the executive is very limited in its powers.
The more power that can be devolved to the state and local level, the better - there's no reason to think a small group of people in Washington are capable of making considered decisions on behalf of 330M+ Americans in the majority of circumstances, and that extends to the myriad of federal agencies engaged in the rule-making process.
This is happening in an election year, so the office of the presidency is driving this for political reasons. I sincerely doubt you'd be comfortable having someone you disagree with politically wield the same powers in the same way.
The presidency is not an elected autocracy. The extent of his powers are strictly limited to those granted by a) the constitution, and b) acts of congress.
Attempts to circumvent these limitations through clever legal theories are undemocratic, doubly so when that circumvention bypasses duly elected state governments. No act of congress has ever explicitly banned non-compete agreements or authorized the FTC to do so, and the plethora of employment law at the state level strongly supports the notion that it wasn't even a federal matter to begin with.
Would you argue that it's more democratic or less democratic when powers previously belonging to states are subsumed by the executive in this way? What if the people of Texas believe non-compete agreements are important? Why not just federalize all laws and tell state legislatures to pound sand?
The goal is to get people to think "Yeah! Taking action on non-competes is great! Darn politicized courts!", when in reality this is not something an executive agency should be doing without an act of congress, or it may not even be a matter that falls to the federal government at all.
Most employment law, for example, falls to the state in which the worker lives, and some have chosen to ban non-competes via legislation. This is much more democratic than attempting to craft law by executive fiat, even though I tend to agree that non-competes are more harmful than good in many situations.
Incredibly hand-wavy. Finance is a way to allocate scarce resources - you can't claim the "technology" is solved if you haven't figured out a way to do it cost-effectively at scale.
No such cost-effective technology yet exists for nitrogen fixation and steelmaking at scale, regardless of what small pilot projects might have been attempted thus far.
"We solved the equations, now it's up to the eggheads to figure out how to finance it" is something only an incredibly naive engineer could possibly think.
> Primary energy demand is going to shrink, not grow.
What? Absolutely nobody is predicting a shrinking primary energy demand. You can hand-wave and say "yeah but electrification", but nobody is predicting a decline in global demand before 2050, if at all this century. I'd argue that demand will more than make up for the efficiency gains as we've seen in the past - look at the growth in deployment of electricity-hungry GPUs as one example of robust demand growth.
It also doesn't matter if US demand shrinks if China and India more than make up the difference. The planet doesn't care what country CO2 emissions come from.
> The truth is that the transition is happening
Who said otherwise? Nobody is arguing that renewables aren't growing.
Vaclav Smil's argument is that historical evidence strongly suggests that primary energy transitions take a lot longer than people want to think, and even as the renewable share of energy is growing the absolute demand for fossil fuels is still increasing. Again, the planet doesn't care about relative share.
People who believe the future will somehow be different from the past (like you) need to provide extraordinary evidence for why they believe this will be the case, and there is none. Installing solar panels, or wind turbines, or upgrading distribution lines, or selling electric vehicles all do not follow any kind of Moores law-like curve, so what factors would drive future results to be different?
Maybe robots will install solar panels faster than humans? Like what's the thesis here?
You can live in a fantasy world and cling to wishful thinking, but it will be a great recipe for mass disillusionment when activists are selling a vision of the future that is at odds with reality.
Vaclav Smil's excellent analysis on this is worth reading: https://privatebank.jpmorgan.com/content/dam/jpm-wm-aem/glob...
The tl;dr is that net zero by 2050 would require all major economies to spend 15-20% of GDP for the next 26 years uninterrupted. For reference, the entire US federal budget was around 23.7% of GDP in 2023.
It simply is not going to happen, and pretending it will greatly undermines the pragmatic conversations we should be having about adaptation.
I think there's a strong optimistic case that the private sector can get us there by 2100 or so - lots of fundamental advancements can happen in that timeframe - but hamfisted government policy in the EU and America that blunts economic growth will mean there's less money to spend on solving these problems in the future.
They have various conservation strategies to manage peak demand if it exceeds capacity, like asking for voluntary conservation and/or rolling blackouts in extreme cases, but every grid operator does contingency planning like this. The Texas grid handled historic demand levels over the summer just fine.
I've been living in Texas since 2022 and the grid has anecdotally been just as reliable as anywhere else I've lived. Unusually crazy winter weather knocks out grids all over - back in '98 the ice storm in the Northeast US & Canada knocked out power for millions of people for up to 2 weeks in some cases.