[aksophist]

I read up to where he started taking questions (less than half the transcript or so?) and these were the interesting quotes that stood out to me:

So imagine a non-arrogant programmer that actually does what you want and you don't have to pay all that money to and there's infinite supply of these programs. That's all within the next year or two.

…

Google decided that work life balance and going home early and working from home was more important than winning.

…

But certainly in your lifetimes, the battle between the US and China for knowledge supremacy is going to be the big fight.

…

And one of the things to know about war is that the offense always has the advantage because you can always overwhelm the defensive systems. And so you're better off as a strategy of national defense to have a very strong offense that you can use if you need to.

And the systems that I and others are building will do that. Because of the way the system works, I am now a licensed arms dealer, a computer scientist, businessman, and an arms dealer. Is that a progression? I don't know. I do not recommend this in your group.

…

And if anyone knows Marjorie Taylor Greene, I would encourage you to delete her from your contact list because she's the one, a single individual is blocking the provision of some number of billions of dollars to save an important democracy.

[HeatrayEnjoyer]

Half of these are comic book villain levels of evil they're so brazen. This person shouldn't be trusted to be in charge of a corner deli.

[latexr]

Even modern cartoon villains are better people than that. Sure, they want to take over the world, but they do so overtly (and wouldn’t try to suppress an interview where they said so) and have a modicum of concern for their own family and underlings.

[ks2048]

Kissinger's buddy Schmidt is in the military AI drone business and drooling over massive cold and/or hot war with China. Not a good sign for the future of humanity.

[chii]

if there was going to be a gold rush, the surefire way to make money is to sell shovels.

[xdennis]

I don't think calling him "Kissinger's buddy" works as a smear since Kissinger was very friendly with China.

I'm not trying to defend Schmidt, but you're naive if you think the US shouldn't develop AI weapons.

How would the world look like now if the US took the high road and refused to develop nuclear weapons and instead let Nazi Germany and Soviet Russia develop them instead?

[datavirtue]

The military had AI deployed on the battle field long before the masses awoke to it via chatGPT. The controversy over using remote weapons and AI was kicked off back in the Obama administration. Before the debate has any time to congeal someone is going to deploy autonomous weapons. Iran may already be doing it.

[danschuller]

> imagine a non-arrogant programmer that actually does what you want

I don't think this is going to be reality anytime soon. In order for the LLM or agent to do what you want, you'd need to be able to precisely specify what you want and that's a hard problem all on it's own. And if you were able to do that precise specification you would be the programmer.

Not say the software developer paradigm won't change but it seems very unlikely to become "make me a better google ads system" anytime soon. I could see getting to something were you are given a result by an agent and then can iterate on it, towards some solution.

[chii]

> In order for the LLM or agent to do what you want, you'd need to be able to precisely specify what you want

no, you just need to vaguely know what you want, and get the LLM to produce something that you then examine, and crawl towards the end goal.

LLM's could potentially allow fast iteration from a laymen's description of what they want.

[xen0]

Sometimes when writing 'fiddly' code, I'll have a bug.

But I can't find the bug. I get the wrong answers but can't trace it through the logic.

Maybe it's a dumb thing like a missing index increment? Or a missing assignment and I just can't see it.

Maybe it's easier to just tear down the mess and write it again.

This is how I feel whenever I deal with AI generated code.

[ljf]

Exactly this - show me something and I can tell the AI what I don't like or what it is missing.

Equally, you can ask the GenAI to keep asking you questions to broaden its knowledge of the problem you are solving, and also ask it to research the issues customers are having with a current solution.

Some engineers seem to imagine any non coder using AI will behave very simply 'make me a new search engine' . Lots of very clever people (who just don't know how to or want to learn to code) will be picking up the skills to use AI as it gets better and better.

I can see AI being used to write far better requirements and produce amazing prototypes - but if you work at a megacorp, chances are (for now) they will want that code rewritten by a 'human' developer.

[datavirtue]

Today, millions of lines of chatGPT generated code will be committed in large organizations.

[ljf]

True, I'm no doubt being too cautious - where I work it is mainly used for unit testing and prototyping - as I understand it we are using it with developers, but always with a human review - we never have a product owner making code with a AI tool and deploying to live. Yet.

[datavirtue]

Oh dear God, if you think product owners are going to be committing code you must be missing something. The AI is to help devs work easier, not replace them. We are nowhere near that. Too much involved.

[djhn]

In what languages can chatGPT write anything remotely sane and deployable?

[datavirtue]

I have had it build out entire API in node Express. Shit needed help for sure. The point of AI is to have it so low value work like scaffolding and boilerplate. The AI goes off the rails a lot. You have to have the skills to recognize when it is and somehow change course.

[pas]

it doesn't matter, it's still likely true. many programmers use it, many of those generated lines will in fact become part of commits. (maybe not millions, but ... it depends on the definition of large orgs.)

[rwmj]

The problem with this plan is reading code is the hardest part of coding. Especially code you haven't written.

[janalsncm]

The issue with LLM driven development is that it’s often as hard to verify the outputs of the model as it would’ve been to write it myself. It’s basically the programming equivalent of a Gish gallop.

[myworkinisgood]

You could also do the same thing with a high-level language. Your LLM is nothing more than an interactive optimizer.

[chii]

but now you got to go learn that high level language, rather than use natural language you already know.

[Jensson]

But now you need to learn how the LLM understands your natural language words, which is very context dependent and will change in the next LLM update. I don't think that takes less time, at least if you are going to write something non-trivial.

[pzo]

Thats still gonna be a big change shift if such companies could axe every 2nd or 3rd developer in their teams. In that situation you might be competing with your colleague not to loose job or have to be "non-arogant" (/s) to ask for pay rise.

[campers]

Another way to look at it is everyone's productivity will be expected to increase to match the productivity increases by the competition who are also using AI. If you don't skill up on how to effectively use AI, then we'll find someone else who has.

[CaptainFever]

This is the lump of labour fallacy. There's always more work to be done.

[HuangYuSan]

Even if you formally specify what you want on a high level and the LLM implements it on a low level, yes, you can call yourself the programmer and the LLM would be a compiler but it would still be amazingly useful

[campers]

There's definitely room to build specification builder agents, that have access to documentation and previous specifications.

The other day I was looking into adding Trusted Types in the Content-Security-Policy header, which was something new to me. In my chat with Claude I asked:

"Lets brainstorm 10 a list of ideas closely related to this so we can think of anything we might be missing on the topic to consider."

And that provided a good list of items to review to consider and expand out the sphere of thinking for the LLM.

It is an infuriatingly hard problem to have the LLM produce excellent results every single time, and have it just do everything and want it to read our mind and all the knowledge and context of a task. I think we'll make some good progress over the next few years as agentic workflows are built out to mimic out thought processes, and the cost/capability of the LLMs keeps improving.

[thaumasiotes]

> And one of the things to know about war is that the offense always has the advantage because you can always overwhelm the defensive systems.

This is not actually one of the things to know about war. It's about as far from the truth as you can get.

[lucianbr]

It's funny how this guy, as most rich business types, thinks he actually knows everything, and he's an expert in all domains. What the hell does he know about war? He's not a general or admiral or something. Of course he imagines being CEO of a large company is the same thing, but it's really not.

Something to keep in mind when reading his other pronouncements.

[prmoustache]

Even successful businessmen don't even know that much about building and running a business as so few of them are able to replicate a prior success and end up pursuing a career of "passive investors".

[lucianbr]

I hate the term "serial entrepreneur". You don't call someone who makes one breakfast "chef", and if they make several breakfasts "serial chef". If you are qualified for something, you can do it any number of times.

Betrays the reality of them being "serial lottery winners".

[pzo]

Palmer Luckey got quite successful with Anduril even though he wasn't that much a business man in the past, but was successful as engineer [0]

I think war in ukraine shows that hord of very cheap sea drones can even thread a fleet of warships.

[0] https://www.latimes.com/business/story/2024-08-09/weapons-st...

[sulandor]

saturation attacks are mostly cheap and effective. see ddos vs intricate hacking

[mbivert]

Indeed: how to avoid needless conflicts is the key. Having a strong offensive is temporary, and potentially difficult to maintain in the long run.

The use of AI/robots in war is probably not a good way to ingrain into people how to avoid conflicts either: as sad as it sound, let them get a taste of it first hand, and that'll probably calm things down more efficiently: not for a few hours/days, but decades.

[notinmykernel]

>> And one of the things to know about war is that the offense always has the advantage because you can always overwhelm the defensive systems.

> This is not actually one of the things to know about war. It's about as far from the truth as you can get.

... Unless you are writing the rules of war:

"Secretary of Defense Ash Carter appointed Schmidt as chairman of the DoD Innovation Advisory Board announced March 2, 2016. It will be modeled like the Defense Business Board and will facilitate the Pentagon at becoming more innovative and adaptive.

In August 2020, Schmidt launched the podcast Reimagine with Eric Schmidt.[71][72] In December 2021, Schmidt joined Chainlink Labs as a strategic advisor.[73] In October 2022, he co-authored a piece titled "America Could Lose the Tech Contest With China" for Foreign Affairs with Ylli Bajraktari, former executive director of the U.S. National Security Commission on Artificial Intelligence.[74] In March 2023, Schmidt testified at a U.S. House Committee on Oversight and Accountability hearing regarding AI.

In 2022, Schmidt was appointed to the National Security Commission on Emerging Biotechnology, a legislative commission charged with making policy recommendations to Congress and the Executive Branch.[1]"

"Since 2023, Schmidt has been involved in building White Stork, a startup developing suicide attack drones.[2]"

[1] https://en.wikipedia.org/wiki/Eric_Schmidt

[2] https://www.forbes.com/sites/sarahemerson/2024/01/23/eric-sc...

[thaumasiotes]

Your response doesn't actually appear to be related to my comment. Does some part of your comment, or one of your links, support the idea that the attacker in a war always has an advantage over the defender?

Remember when Georgia invaded Russia?

[aguaviva]

Remember when Georgia invaded Russia?

We don't actually. Care to elaborate?

[matthewmacleod]

If only there was some kind of ongoing event that demonstrated exactly this…

[adastra22]

Or, you know, a whole world war that definitively proved the capability of dug-in defenders to hold ground.

[lesuorac]

I think the second world war showed how big the attacker advantage is.

Germany got's industrial base absolutely shattered while the non-defending US was able to use its industrial base to supply arms for 4 fronts (vs Germany, in Africa, for Russia, vs Japan).

[kolme]

> And one of the things to know about war is that the offense always has the advantage

I'm no military expert, but wasn't it exactly the opposite?

The attacking army needs to be bigger than the defending army and will suffer way more casualties,even if they are successful.

[fullshark]

The premise is that the offense knows the battle plan beforehand and the defense doesn't I believe, one of those "be aggressive" executive metaphors I think he internalized but not the reality of ground warfare, since probably the US Civil War when trench warfare took over at the latest.

[borski]

Not necessarily. The Trojan Horse wasn’t so disastrous because it was filled with way more soldiers. It was because they had penetrated all of the defenses and had the element of surprise.

Iran just sent I-don’t-know-how-many drones and missiles at Israel like a few months ago. A few of them landed, but most were caught and intercepted in the air. Here’s the thing: if even one of them had hit, say, the center of Tel Aviv, or the old city in Jerusalem, it would have been a massively successful attack, even if none of the other ones had done any damage. The size of zero armies was measured in that exchange.

[latexr]

The Trojan Horse is a myth. It’s as meaningful an example of military strategy as Gundam robots.

But even if those were real they would still not support Schmidt’s point that “offense always has the advantage because you can always overwhelm the defensive systems”. The Trojan Horse didn’t overwhelm defenses, it penetrated them and destroyed the enemy from inside.

[borski]

Sure, that's fair - I don't mean to imply that it is always true, and 'overwhelm the defensive systems' isn't the language I'd use. All I mean is that specifically targeted strikes, at the right targets, at the right time, can sometimes be far more important than who has the bigger army. Sometimes if you cut off the head, the rest of the snake really does kind of just die.

[meiraleal]

I guess the meaning is that defending already starts losing. It is much better to be in the offensive, it takes your enemy much more effort to go from defense to offense

[CRConrad]

Deep strategic thinker and tactical expert[1] that I am, I agree: As the old maxim goes, "the best defense is a good offense".

[1]: Having watched the evolution of a football team's tactics from 5-v-5 (2 min, age 6-7) to 11-v-11 (2*45 min, age 18-19). A regular Strategos, me.

[TrapLord_Rhodo]

but you are missing context here. His point was that once cheap $500 drones can be produced at an industrial scale, in a highly automated fashion you can completely overwhelm any defensive measure that could be put in place.

Normal warefare is based on "Manpower" and you reference "Army". He further goes on to explain that artilery, tanks and any kind of "land army" is completely obsolete.

[CRConrad]

The word "explain" works a bit like "know" (and perhaps "teach", I think). No, you don't "know" that the Earth is flat; you just mistakenly think so. (And no, you don't "teach" kids that the Earth is flat; you just mislead them.)

To say that Schmidt "explained" that artillery, tanks and armies are obsolete implies that that is true, which... At the very least, remains to be seen. The correct term would be something like "He further goes on to bloviate about how artillery, tanks and..."

[bjornsing]

But when those “casualties” are drones perhaps it turns out to be true?

[tetha]

It might be a reference to the blue-team red-team asymmetry, and how in cybersecurity the attacker has an advantage. The attacker there only needs one success and can rapidly try different avenues, while the blue team just needs to miss patching one system and that's it. And while patching may be technically simple, the organizational efforts around it are sometimes... eh.

In war, defenders can entrench more and more and a lot of work and planning is put into hitting either before the defenses are up, or not at all.

[meiraleal]

> So imagine a non-arrogant programmer

Imagine a non-arrogant CEO

[ImHereToVote]

"democracy"

[refurb]

Damn, the Marjorie Taylor Greene comment sounds like “this person is bad because she isn’t interested in the US getting involved in more foreign wars”.

That’s actually an endorsement .

[weweersdfsd]

There's nothing wrong with getting involved in foreign wars, if that means supporting a nation that has been attacked by its neighbor in gross violation of international law.

I don't agree with US involvement in Iraq, Afghanistan and many other places, but Ukraine is a obviously totally different from those. Marjorie Taylor Greene may not give a damn as she lives thousands of kilometers from Russia (or maybe she's just been bought by Russian money), but morally she's wrong. When it comes to Russian imperialism, the only right thing to do is help its victims fight back.

[floydnoel]

there is everything wrong with getting involved in foreign wars. maybe you should join the military and participate in a war before thinking they are so great.

i've never agreed with MTG before and i think she's a terrible person but i applaud her stance in this case.

[weweersdfsd]

Well, try living in a country that shares over a thousand kilometers of border with Russia, and which has been invaded many times by them during past 300 years, resulting in massacres of civilian population. Then you might see things differently.

Without US playing world police my home would probably have been bombed by now... Or maybe Hitler would have won and destroyed Russia, who knows. In any case, I much prefer living in an independent country protected by NATO, over being a member of an enslaved ethnic minority inside Russia, in which case I would have probably been sent to the front lines in Ukraine already.

A flawed world police is better than no world police at all.