A model that gets good at computer use can be plugged in anywhere you have a human. A model that gets good at API use cannot. From the standpoint of diffusion into the economy/labor market, computer use is much higher value.
I can honestly understand both positions. The U.S. military must be able to use technology as it sees fit; it cannot allow private companies to control the use of military equipment. Anthropic must prevent a future where AIs make autonomous life and death decisions without humans in the loop. Living in that future is completely untenable.
What I don’t understand is why the two parties couldn’t reach agreement. Surely autonomous murderous robots is something U.S. government has interest in preventing.
Your comment reminds me of a story. John Adams and Lafayette met in Massachusetts something like ~49 years after the revolution. (Lafayette went on a US tour to celebrate the upcoming 50 year anniversary of independence.) Supposedly after the meeting Adams said "this was not the Lafayette I knew" and Lafayette said "this was not the Adams I knew".
Yeah, it's the easiest way to get a beachhead in deep learning and then expand from there. I dislike their heavy use of currying, it's elegant in theory but bad error messages make it confusing and inconvenient in practice. But it's a small tradeoff for an otherwise excellent book.
Ah, I'm using racketmode which doesn't support live state buildup (and the builtin GUI doesn't either). What exactly is your setup? SLIME only has a Common Lisp backend, it doesn't support Racket to my knowledge.
EDIT: ok with geiser and geiser-racket incremental state buildup works really well. I rescind my objection!
I've been using Racket to work through The Little Learner[1] and it's been a good experience. You need minimal Racket to work through the book (lambda, let, define, map; I think that's about it). But I branched out to learn more about the language and the standard library, and it's a fun and surprisingly powerful system to explore.
The biggest downside of Racket is that you can't build up your environment incrementally the way you can with Common Lisp/Sly. When you change anything in your source you reload REPL state from scratch. After CL it feels incredibly limiting in a Lisp. Incremental buildup is so valuable, if I wanted to do any Lisp work again I'd reach for CL before Racket just for this reason.
BTW, the book is _great_. Quick, easy to get through, very easy to understand, and teaches you everything from soup to nuts. If you're familiar with lisps you can get through the book in two weeks. It's then easy to get into any deep learning tutorial or project you want, or even start implementing some papers. The book manages not to water down the material despite not using any math at all. Although if you know some linear algebra or multivariable calculus you'll appreciate the beauty of the field more.
Also ex-Stripe. This suggests an opportunity to build an exchange that addresses these problems. Could one build an exchange with deliberate "turn-based" liquidity to avoid the problem of daily stock price distraction, for example? (This is hard because there will always be secondary markets, but presumably this is already the case.)
I thought about it for only a few seconds, but here is one way to do it. Have users self-report an "addiction factor", then fine the company based on the aggregate score using a progressive scale.
There is obviously a lot of detail to work out here-- which specific question do you ask users, who administers the survey, what function do you use scale the fines, etc. But this would force the companies to pay for the addiction externality without prescribing any specific feature changes they'd need to make.
From an execution standpoint you can't work on experimental mobility due to path dependence. How are they going to convince municipal governments to open golf cart lanes? That would require solving two problems (autonomy and overcoming path dependence), and solving just one is hard enough. Once they saturate the market as it is with autonomous driving, then everything will change and opportunities to experiment will open up.
To be intellectually honest about it, you have to answer a bunch of questions:
1. Awful compared to what?
2. Was there an equivalent transfer outside America?
3. What is the cause? What ratio rent-seeking/shady activity vs a consequence of natural forces (e.g. technological change)
Azure revenue is growing at 39% year over year. If Microsoft can sustain this growth, in four years Azure will be ~3.73x its current size. This is of course very difficult, but you really don’t need a deus ex machina to hit 4x growth under your assumptions.
I've recently switched to Claude for chat. GPT 5.2 feels very engagement-maxxed for me, like I'm reading a bad LinkedIn post. Claude does a tiny bit of this too, but an order of magnitude less in my experience. I never thought I'd switch from ChatGPT, but there is only so much "here's the brutal truth, it's not x it's y" I can take.
The vertical integration argument should apply to Grok. They have Tesla driving data (probably much more data than Waymo), Twitter data, plus Tesla/SpaceX manufacturing data. When/if Optimus starts on the production line, they'll have that data too. You could argue they haven't figured out how to take advantage of it, but the potential is definitely there.
I work for Microsoft/Azure and my incentives are (roughly in descending order): minimize large/long outages, ship lots of stuff (with some concern for customer utility, but not too much), don't get yelled at for missing mandated work (security, compliance, etc.) I'd love to improve product quality, but incentives for that are negative. We're running a tight ship, and every second I spend on quality is a second I don't spend on the priorities above. Since there isn't any slack in the system, that means my performance assessment will drop, which I obviously don't want. Multiply that by 200k employees, and you get the current state of quality across the whole product portfolio.
I still don’t understand why people oppose that rather than enthusiastically desire it. The end state you’re describing is the culmination of the enlightenment project. Automating labor is the point! Then you can paint, or play chess, or eat amazing food, or do whatever you want. Work isn’t the end, it’s the means. Products and services is the end. If we can achieve the end via technology, who cares about the work?
That America works and keeps working for reasons nobody understands. We seem to make a ton of mistakes, and often act in the same way as far less successful countries do, but somehow keep coming out on top. It often feels like our success is despite our actions, not because of them. That's what I mean by "intangible American magic".
The surface level reading of history is that everyone always agreed centralization of power is bad and operated in a gentlemanly way to further this principle, until Trump came along. But that's not the case at all. In the two parts of American history where I read more or less deeply (early republic and civil war), the factions bitterly vied for absolute power but were prevented from getting it through intangible American magic.
From my reading of history the level of polarization was at least as high or higher during the early republic, and obviously higher leading up to and through the civil war. I don't know American history in any meaningful depth outside these two periods, but I suspect there were other periods of extreme polarization. I really don't think this is new.
In the 1800 election republicans thought the federalists will turn America into a tyrannical monarchy, and the federalists thought the republicans will plunge America into mob rule like the French Revolution. They would have never defended each other.
Things look bad if you have a twenty year time horizon, but they look pretty normal if you zoom out to encompass all of American history.
Re Trump exacting revenge on political opponents, that conduct has endless precedent in American history. (Refusing to concede the election does not; but he was forced out of office nonetheless, which I read as a sign the republic is healthy.)
EDIT: I just remembered about the Aaron Burr conspiracy. Aaron Burr lost the 1800 election to Jefferson by one elector (after over a month of gridlock). He then tried to raise a private army to either march on Washington, or to form his own country and secede from the union. In 1805 Jefferson ordered his arrest and Burr was tried for treason (and acquitted!)
Yes, but this tension has been there since the founding of the republic. The federalists wanted to centralize power (some much more than others), and the republicans (the "jeffersonians") bitterly opposed it. In his second term George Washington personally lead troops to Pennsylvania to put down the whiskey rebellion.
Zooming out of the 24h news cycle, "all power to a singular person" concerns seem far too overblown. Half the country hates Trump. He won the popular vote, but not by all that much (despite what he may assert). By comparison LBJ, FDR, and Nixon won ~60% of popular vote. Even if he were a young man, I don't think we're in any danger of a caesar.
What is the evidence that America is falling apart? From all my reading of American history, America has _always_ been this way. With a wide lens it appears as healthy as it's ever been. This is a genuine question-- I've read a lot of American history, but I'm still a dilettante. It's extremely difficult to tell if there are genuinely new conditions, or if we're engaging in a vigorous political process as we always did.
This is a half-truth that obscures what makes The United States unique. From the day the constitution was signed, it was a compromise between competing economies, geographic incentives, cultures, religious movements, political philosophies, and individual ambitions. This is what makes the country free and prosperous-- it was conceived under conditions that make centralization of power extremely difficult.
I know you didn't ask to problem solve, but... Email your resume to me at vakhmechet@microsoft.com, mention this HN thread. We're hiring for Azure MySQL and sister Postgres team, lots of interesting problems to solve!
EDIT:
"I feel like a firework that exploded in bursts of color (everyone ooed and ah-ed), and then... nothing" ~~> that's a beautiful sentence. Welcome to adulthood!
My experience has been that "chosen family" is a thing that works when you're young, but almost always falls apart when you get older. This has happened to countless people I know. Life throws all kinds of curveballs, incentives change, conflicts arise, sometimes very intense conflicts. Empirically, chosen family is a structure that works in a particular place and time, then disintegrates when conditions change. Real family isn't like that; there is a very strong anthropological connection wired into us that doesn't go away when the situation changes.
Of course it's different for everyone, some families are so tragic they may not be worth preserving, etc. But that's an outlier-- the modal experience is that the power of family is precisely in the fact that you don't get to choose it.