Andreas Brakkerud

I like places where nothing much happens.

[Benjamin Bratton @bratton]

Probably a good time to re-up Agentworld, antikythera's new research unit launched earlier this summer. Our white paper/ call for papers anticipated many of the current debates on agent civilizations, anthropomorphism, agent institutions, agent phenomenology, etc. See https://t.co/aqJynGZyIt

[Matthew Fraser 🇪🇺 @frasermatthew]

Younger people may not remember Edouard Balladur, the man who almost became French president. Three things come to mind when I recall his time in the political spotlight in the late 80s and early 90s. https://t.co/urdyCXdtFx

[Chaz Wargnier ♨️ @DataChaz]

In Pakistan, a man crashed into a police car while testing a self-built driverless car.

@TESLA, PLEASE H-1B THIS GUY ASAP

[liemandt @jliemandt]

Today, I’m honored to join Andrew Huberman for a deep conversation about how children learn and how we can redesign school around what science makes possible.

Just one year ago, I recorded my first podcast with @patrick_oshag on Invest Like the Best.

Since then, word of mouth and demand for Alpha School have exploded.

We’re now bringing Alpha to 50 communities across the country, each made possible by founding families determined to build a better education for their children.

[Andreas Brakkerud @andersbrakkerud]

🇳🇴😢

[Bruno Maçães @MacaesBruno]

Portugal, Spain, Poland growing faster than the US. If only Germany and France got their act together. Stop blaming foreigners and reform your economies like Portugal and Spain and Poland did fifteen years ago

[Andreas Brakkerud @andersbrakkerud]

Lol

[Bruno Maçães @MacaesBruno]

We now have a perverse situation where AI companies have an incentive to produce malfunctioning software because they can then present it as having a mind of its own. Different to when a company made automobiles or airplanes

[Bruno Maçães @MacaesBruno]

Biggest non-surprise of the year

[Andreas Brakkerud @andersbrakkerud]

As predicted by Napier et al

[Financial Times @FT]

Risk of a new age of financial repression is rising https://t.co/nYLpUyoDLq

[Jordan Weissmann @JHWeissmann]

Did the fall of Rome feel this constantly stupid?

[Sar Haribhakti @sarthakgh]

The vast majority of personal productivity stuff (emails, calendars, tasks, crm etc) people share as examples of things they are doing with personal agents doesn’t resonate with me

It all sounds very boring, unnecessary and uninspiring

[Santiago Capital @SantiagoAuFund]

another shocking development

[Hollis Robbins @anecdotal]

In which I suggest that a 1930s dictionary-based personality test is perhaps not the best instrument to try to give a large language model a personality. It's worse than circular (and bad social science). https://t.co/IGbm7X80by

[Branko Milanovic @BrankoMilan]

My today's Substack:
On European hypocrisy and warmongering
https://t.co/kylPAZikBT

[Gérard Araud @GerardAraud]

L’histoire nous apprend que vient le désastre le moment où la majorité ne comprend pas que les extrêmes entendent réellement abattre le système en place et mettre en œuvre leur programme.

[olexander scherba🇺🇦 @olex_scherba]

With all the horrors of this war, this is particularly evil.

🇺🇦 Donbas. A Russian drone operator spots a young civilian on a bicycle, looks him in the eye and kills him in cold blood. And posts it on social media.

[T. Greer @Scholars_Stage]

I am going to make a point that seems to be missed in all of this. Better messaging—ie, never talking about AI taking jobs—is not the problem. The problem is that AI just has not delivered too many tangible benefits to the country writ large. Consider the cutting edge accomplishments of the cutting edge models. The frontier has been pushed in the world of pure mathematics and in agentic coding. But what good has come of that? The main effect of the first: supplanting grad students from their budding careers in mathematics. The main effect of the second: rogue LLMs hacking other companies.

People keep tweeting things like “being against data centers is being against cancer research.” But why would anyone think this? Are any of the frontier companies focusing their resources on training their models to solve cancer? Or to engineer a breakthrough in materials scienxe that might allow us to lower energy costs? To provide any scientific or engineering breakthrough that might directly, obviously benefit normal people?

The average person thinks of LLMs as the thing their kids use to cheat on school assignments. AI may have made their life more convenient on the margins but that is the margins. So you end up asking communities to accept these data centers and sanction the local tentacle of a technology that is already making people far away very rich but which has not provided the country writ large wirh any tangible goods.

In other words: the problem is the substance not the messaging.

[kache @yacineMTB]

The universal decompiler is here. It is the worst it's ever going to be. It is only going to get cheaper and faster

[Andreas Brakkerud @andersbrakkerud]

@danielbelfort Good point :)

[Andreas Brakkerud @andersbrakkerud]

Half an hour on Claude code or Hermes gets you the same, but why not pay extra for a wrapper

[Daniel Belfort ⊙ @danielbelfort]

Introducing Offprint™

The physical magazine made from everything you "save for later" but never actually go back to read.

Find online, read offline.

Print media is back!

[Andrey Mir @Andrey4Mir]

For ancient humans, nature was the clock: day-night, seasonal change, biological cycles. The environment dictated what to do. But the more we used media, the less nature dictated how we spent our time. Fire extended daylight; technologies overcame seasons (and even biology).
1/