rockskon

Bait and switch user interfaces to insert AI into places AI did not previously exist and make sure you see an AI answer before anything else is not far divorced from the "lead in food" example - especially when paired with efforts to make the 10 Blue Links become irrelevant garbage on average by ignoring most of the words you type in.

Maybe it's a better analogy to compare this to food-without-lead costing 10 times as much as food-with-lead in large part because of direct actions undertaken by lead manufacturers.

That worked so well for the finance system finding new Cobol programmers!

Because eventually you'll get to the point where you've too much work to do and there's not enough people to delegate it to.

Hope you like being overworked!

Oh good lord not that statistic again.

Left unstated is what jobs philosophy and art history majors take.

There's more computer scientists working in computer science than there are philosophy or art history majors working in philosophy or art history.

That various companies such as Google are working to kill. They're an advertising company that is making it increasingly clear they no longer want to link to their competition. Competition being defined as any source of information that is not Google.

Has it?

Spam calls frequently don't have a source in the same country as their target victim.

No.

Fair.

I don't have a good frame of reference to know when VCs consider the industry worth pouring more money into or if it's a black hole for investment.

Surely there's gotta be some point where the industry has to prove itself worth additional investment $ rather than banking on finding another company where over 80% its valuation is based on memes and dreams (see: SpaceX) or has cartoonishly inflated expectation of future profitability (see: Palantir).

Have many defense investments in startups proven profitable enough for VCs and other investors to continue pouring money into the sector? Or is the sector primarily banking on moonshots?

I'm aware of a relative lessening of cultural aversion in the bay area of having anything to do with the DoD, but what of substance has changed on the gov side, really?

I'm aware of DIU and of greater outreach efforts to the bay area. I'm aware of a greater rhetorical focus within DoD leadership on taking advantage of Silicon Valley's tech expertise.

But I'm not aware of any substantive change to DoD acquisition law to address the fundamental incompatibility in funding models between DoD's traditional contractor base and bay area startups. The significant lag time to award contracts. The risk-aversion for the DoD where civil servants face political and, in some cases, legal liability for granting funding for projects that don't pan out.

Has the DoD's risk tolerance for investments changed recently? As in - accepting the possibility that some investments may not pan out if it means there's a good chance that other investments bear fruit.

Last I checked a few years ago, outreach to the bay area startups has been primarily using the DoD's equivalent of change found in the couch. A rounding error of DoD's overall available funding.

> This isn’t true. They literally do this all the time. They just need funding. This is also true for biotech.

Thus the sentence I immediately followed the one this was made in response to where I said "They tend to die in that time if they have no funding."

> Right. So they raise funding.

In many, many cases when it comes to the DoD, their wants aren't seen as dual-purpose and start-ups struggle to find funding that isn't from some DoD-aligned and defense-focused investment firm - which haven't historically invested in large numbers of startups. At least not when I last checked several years ago.

And just to get ahead of this - a DoD want not being seen as dual purpose and the tech later being used for a dual purpose are two very different things.

> Your argument boils down to “the DoD won’t work with startups that don’t have funding,” which is both true and, frankly, as it should be, in my opinion.

My argument is that DoD contract law is poorly suited for funding meaningful sums of money to start-ups that do not have significant non-DoD sources of funding. I'm to understand relatively small sums of money can be awarded on a short time scale, but those sums of money are tiny compared to what's needed to execute on most contracts.

I primarily alluded to AI because of obvious reasons, but yes - Cloud computing and related service products are also good examples.

But there's a great deal of department needs that don't revolve around information processing infrastructure where there aren't many success stories to refer to. Especially relating to startups.

I question what specifically you're referring to.

Yes, Hack for Defense is a decade old now. But the DoD famously had not done much business with area startups for many decades outside of very specific success stories like the CIA's In-Q-Tel.

Turns out that start-ups can't wait several years for a contract award. They tend to die in that time if they have no funding.

Additionally - talk of electrical engineering work done for world war II / cold war radar technology has been a oft-repeated tagline by members of military leadership as well as Palantir representatives when talking amongst themselves about Silicon Valley or in their appeals to SV itself.

"We have so much in common! Here, why don't you open your history book and I'll show you!" - that's what the appeal comes off like.

I maintain that primarily relying on those examples is a poor choice in trying to establish cultural similarities.

A tiny number of enormous companies capturing the department revolving around one specific technology doesn't sound like it's bridged the acquisitions gap between Silicon Valley and the military.

That's about as culturally relevant to Stanford as talking of people who lived through the great depression.

The DoD doesn't get to neglect relationships with a community for decades and then talk of how much in common they have with each other. It's nonsense and transparently manipulative.

Well - it's not exactly a surprise that all these non-American countries engage in un-American practices.

It's much more concerning when said practices are undertaken by the U.S.

Just because other countries do something isn't a justification to bring the practice into the U.S. despite that being a justification used with increasing prevalence these days.

Does the average person think this? Perception of what other people think doesn't always line up with what they actually think.

HN has its own share of enticement/manipulation - most recently with regards to AI.

Plenty of variants of "people who don't use AI will be left behind" are sprinkled around various threads about AI. It's an attempt to both manipulate via fear as well as sell.

Don't underestimate the capacity for the problem to get significantly worse.

I do question at what point AI could be useful as a teaching aid.

The quality of LLMs depends heavily on, among other things, how you word your questions.

Knowing the correct questions to ask is not something most students know how to do given that it tends to require a fair bit of pre-existing domain knowledge.

Of course - it's SEO's fault that Google search frequently ignores words in your queries.

It's SEO's fault that Google frequently prioritizes giving you results that are semantically or thematically related but ultimately irrelevant to what you searched for.

It's SEO's fault that Google refuses to place ads on websites not full of meandering irrelevant "content".

It's SEO's fault Google search all-but-ignores most single-purpose tool websites.

It's not the result of decisions made by Google's programmers and project managers - it's allllll the fault of SEO!

The SEO industry definitely isn't an all-purpose scapegoat for horribly unpopular decisions made by human beings at Google.

*2024, not 2014

I did Google this. It seems to be a small number of banks doing this right now. A far cry from "all banks".

Maybe we'll get there eventually? But you should be more wary about making such absolute claims.

I've found that framing topics like this as primarily a pretext for different motives is a sure-fire way to be ignored by people you may want to convince.

As always, the goal in convincing others is to take someone from their current understanding and bring them closer to yours. You can't get there if you don't start the topic at their current understanding of it.

Trusting untrustworthy companies aside, that doesn't resolve the issue of hacked facial recognition data.

Even back in 2014, malware was coming out that steals facial recognition data directly from smart phones themselves.

https://www.theregister.com/security/2024/02/15/stolen-ios-u...

The GDPR isn't a silver bullet.

Additionally - furtherance of facial recognition technology would impact travelling to foreign jurisdictions.

One of the most common ways foreign travellers get flagged when travelling internationally is for social media posts made under their own name that their destination country's government may not like. Traditionally if you've kept yourself pseudo-anonymous, you've largely been safe. But if we get to a point where pseudo-anonymous accounts are associated with pictures of people's faces, it will become significantly less pleasant to travel internationally for a lot more people.

Do you have proof of this? This comes off as conjecture.

A bank's security camera feed isn't likely to be sold to dozens of companies.

It is all but guaranteed for Internet-based facial recognition services.

Hacked facial recognition data already is reportedly being used by scammers to not only bypass bank security but also to impersonate people to target their loved ones.

There is no lasting security gain by providing a selfie. There is a lasting security and privacy loss, however.

Asinine requirement by whatever risk management firm they use. A selfie provides nothing in terms of lasting security while simultaneously adding permanent risk.

Isn't AI routinely making significant mistakes in analyzing medical imaging?

"Their lawyers will win" is plainly wrong assumption if something is clearly illegal.