Aren't the recent results in mathematics actually stronger evidence for his point? Although the models may be capable of generating proofs they aren't coming out with the same level of quality of a human discovered and communicated proof. Providing a gobbledy-gook yet technically correct proof (generated at least in part by brute force) lacks the qualities of an expert produced proof because they fail to communicate insight or understanding about why the theorem is true.
Gaining and successfully communicating insight and understanding from a proof you discovered is additional work that human mathematicians do. It's not just some side-product of proof-finding (at least not to the degree usually needed to publish). That AI models don't provide this is mostly proof that the model wasn't asked to do this work. Either because the prompter didn't know or didn't care
But there are also plenty of examples of humans providing technically correct proofs without any elaboration. Usually they get ignored, unless they are famous or the problem they solved was famous
Coding being the act of encoding/decoding information from one representation to another. The system itself is called a Code and these are designed to have specific properties like making transmission of information resistant to interference, corruption or interception, etc.
In this case, corporate management holding the purse strings but their workers (devs) using the actual tools. The solution they offer to founders is to make the user your champion and have them sell your product for you.
"The meta point here is that you're not going to talk to the credit card holder; the user/dev is going to do that for you.
Give them the best possible chance at convincing the leadership. Make them look awesome for even bothering the leadership with a choice like this. Make it obviously awesome for them to decide “yes”. These users/devs are your sales people."
Maybe that works for dev tools with freemium models, but in many industries where this problem arises its just not possible to even get your product in front of the users. Take hospital systems and EHR purchasing where Doctors and Nurses are the users of the EHR day in and day out but it is the hospital administration that ultimately gets to decide which EHR is deployed. How do you get users to be champions of your product if you can't even get it in front of them?
Too often we get the reverse. Slick salesman targets the person with budgetary discretion while avoiding letting the users in in the transaction, so by the time they can complain about how terrible the product is, the check has already cleared.
Speaking as software developer and company co-founder for a product for use by other developers, the company employing those salesmen is very possibly overall losing money on these deals.
In our early days we had on occasion managed to directly convince non-developer decision makers in an organization to use our product (even though it wasn't an intentional strategy) and this invariably failed unless we also managed to get full buy-in from the developers who will actually use it.
These days we have an explicit rule that we don't go ahead with any customer engagement until we can see that a project's lead developer is fully sold on our product, otherwise we waste loads of time (which is essentially money) on ultimately fruitless retention attempts instead of using the time more productively on ideal customer opportunities.
Now it does depend on the nature of the product though. For a product used by only a small subset of a company (like our product), it's probably a bad strategy to sell to the execs instead of the users. But for a product like an ERP (think SAP, or Oracle ERP), these are deals worth USD millions (sometimes 10s of, or more) and convincing the execs is a highly (or possibly the only) effective strategy.
Yup, then the technical people have to deal with the bullshit and nightmare that is implementing the shitty thing that was purchased. Hard to get around this in large enterprises, someone just decides they're going to use some shitty tool (like that cloud provider no one uses but has great sales folks) and you just get a notice you have to migrate all your apps to the new thing in 3 months cos they got a better contract there.
The school district my kids are changes the parent app almost every year, its always a nightmare for everyone involved, I can't imagine what it is like to work IT in such a place.
I believe open source is approximately an order of magnitude larger than it would be if developers controlled their own purchasing. What FOSS introduced was the ability to use software without someone with a little power saying no, you can’t, because we won’t pay for it.
Jetbrains threaded this needle for years by having professional licenses tied to an individual with clauses for time and location shifting. So you could use their software at home, drive to work and also use the same license there.
And they priced it at around the cost of three tech books per year, which it is at least that useful for productivity. I suspect we would be in better shape now if others had copied their model. Rather than the (defunct) Microsoft model of ignoring home piracy and demanding commercial licenses from any company large enough to make it economical to fire off a cease and desist to them and demand back pay.
> Jetbrains threaded this needle for years by having professional licenses tied to an individual with clauses for time and location shifting. So you could use their software at home, drive to work and also use the same license there.
It's a very, er, "enlightened self-interest" model, because it makes me "sticky" as a customer, since I'm less likely to learn a completely different IDE for work and then use that one for my own projects and eventually ditch theirs.
The less than 400k population is a good sample size, however, their main revenue comes from tourism, aluminum smelting, and fisheries, and they have a relatively large source of geothermal and hydropower with a very homogenous population (>80% Icelandic). Data centers are about 1% of its GDP. Even Iceland needs barkeeps and restaurant workers to work at least 5 days/week with restaurants in touristy areas open 7 days/week with shorter days on Sunday. Having worked a life of manual labor and white-collar jobs, sometimes both in the same job, the 4-day work week is really for specific types of industry or work. Tech, for sure, but not the food service industry...
Rodney Brook's keynote at Stanford HAI earlier this month.
Some interesting points right at the beginning relevant to recent on-stage demos by companies including like Nvidia and Tesla.
- Product lifecycle has a realistic timeline on order of decades going from research lab demo to a realized commercial product. He first saw demos of driving cars arriving in 1979, yet we still don't have Full Self Driving in 2025.
- @2:35 "Humanoids they're everywhere and what is it doing. The form of a humanoid is promising that it's going to be able to do everything that a human can do. And it completely fools VCs. A lot of people's pension funds are going into all these robots and its not going to end so well."
Rodney Brooks states he's built 4,500 humanoid robots in his lab.
Well there are now many instances of enforced disappearances.[1] To what the administration likes to call jails in Ecuador, except for the fact jails and prisons are part of legitimate criminal justice systems with judicial review/due process. These can be more accurately described as concentration camps given that they lack the features that would make them legitimate jails or detention facilities.
"
[Enforced disappearance] is characterized by three cumulative elements (defined in A/HRC/16/48/Add.3):
A) Deprivation of liberty against the will of the person;
B) Involvement of government officials, at least by acquiescence;
C) Refusal to acknowledge the deprivation of liberty or concealment of the fate or whereabouts of the disappeared person.
"
If you think about the administration's unwillingness to comply with the court's ruling to return the individual, who by their own admission, they mistakenly took away due to an "administrative error" there are many open questions. How do we know that the individual is still alive? For that matter, how do we know that all the other people who they say were removed from the country are still alive?
We have no independently verified information as to fates of these people. More likely than not, in the course of these actions by the government, the number of deaths is some number greater than zero. Even if they have not performed outright executions, some deaths as a result of the conditions and or their treatment in custody is almost certain. So is that state sanctioned man slaughter/murder? Does this make ICE a death squad?
What distinguishes a concentration camp from a prison (in the modern sense) is that it functions outside of a judicial system. The prisoners are not indicted or convicted of any crime by judicial process.
There have been and continue to be technological improvements that reduce the dosage required. There is not a single dosage for all CTs. The dose for a head CT is an order of magnitude less than for abdominal/pelvic CT.
"The current average effective dose of a CT study is ∼10mSv, with the implementation of dose reduction techniques discussed herein; it is realistic to expect that the average effective dose may be decreased by 2–3 fold." 2010
https://www.ejradiology.com/article/S0720-048X(10)00311-6/fu...
It would be neat to see a chart of the average effective dose of CT studies over the past 40 years. And any accounting of how much it declined as a result of Moore's law and software improvements for producing the "Computed Tomography."
