The content is corporate. The EU AI Act is extra judicial. You don't have to be in the EU to adopt this very set of "AI Principles", but if you don't, you carry liability.
This corporate crap makes me want to puke. It is a consequence of the forced bureaucracy from European regulations, particularly the EU AI act which is not well thought out and actively adds liability and risk to anyone on the continent touching AI including old school methods such as bank credit scoring systems.
Telegram channels are upstream of ALL text based social media and to an increasing degress, culturally upstream to many content trends as well.
Take the war in Ukraines for example, the uncensored and real time updates you get in open Telegram channels make most intelligence agencies except for Five Eyes nations look regular.
This deal may be much bigger than it seems off the bat. The cohort of people using Telegram to exchange content is maybe the top 5% of the world in many important niches.
In my humble opinion, Google is the reason the web withstood as long as it did. At least the open part of the web did.
Yes, sure, they monetized, but also they gave back as much as, and if not more than they took. We have so many machine learning frameworks, tensorflow, research, payouts to creators, advertising opportunities, careers, products, a lot of things built and taken down but most importantly built. They were probably the most positive force for the internet age in the past 20 years and more than anyone will ever give them credit for. Only in retrospect will we realize how lucky we were to be alive in the Google age. Full stop.
What really killed the web was the rise of closed wall gardens platforms such as Apple, Facebook, Instagram and others. Putting up walls around content that didn't need to necessarily exist or not honoring open frameworks to exchange information and making things more widely indexable.
But even here there have been significant benefits. The present AI boom would arguably not have been as large as it is right now without Mark Zuckerberg choosing to put an unconventional amount of investment behind AR ambitions to take on Apple, an investment the size of which many conventionally run publicly listed or private enterprises could hardly imagine to take up, leading to the concentration of capital, talent, technology and hardware in a place that gave birth to open source Llama and others. Google as well was very well poised because of their investments in compute fueled by their business model which kept the web alive and also returned capital into places where computer scientists would be paid significant amounts of money and have job security and freedom of will. to do as they pleased as opposed to chasing a paycheck, working as a physicist at CERN or something.
All I'm saying is this article does not fully capture how significant the positive outcomes from Google have been.
So have you sat in a Tesla with the latest hardware or . . .
I don't understand this attitude in the technology industry. If you want to hold such a strong opinion on something, at least take the initiative to research what you're talking about.
Teslas __today__ are at or better than Waymo at autonomy. They are launching tests in June. There are popular accounts who have experienced this alpha at the "We, Robot" autonomy event earlier last year and follow on interviews with Lars and Franz, (Head of Vehicle Engineering and Head of Design)
Each Waymo is > $140,000 of customized hardware and is limited to specific cities. Autonomy in commercial vehicles is arguably led by Tesla on coverage, miles driven, ready hardware, cost per mile etc. They’re going to start pushing tests on their consumer fleet, converting them to optionally commercial taxi rides soon with the fleet owner model versus the central provider model. This is scheduled for June in Austin and confirmed to be on schedule.
You can also take fully autonomous bus rides in China right now, even there, for, early reviews, the latest Tesla Autopilot blows everything else out of the water.
I’m not trying to push Tesla alone, but I’m trying to highlight the gap in adoption goals. What is Waymos ambition this year? How much can they ramp their fleet at $140k per unit versus Teslas consumer fleet and upcoming low cost robotaxi with the mass manufacturing improvements further lowering cost per unit?
Not on cars, not in robotics, not in commercially deployed AI, not in enterprise investments in their cloud business.
They've got immense potential, sure. But to say that they're winning is a bit far from reality. Right now, their Cloud AI offerings to the enterprise are technologically superior to anything else out there from AWS, but guess what? AWS seems to have significantly more %age sales growth in this space with their larger base compared to GCP with their smaller market share.
The same can be said across turn based chat and physical AI. OpenAI continues to be the growth leader in the consumer space and a collection of Claude + self hosted + Gemini now in the enterprise / API space.
They need to be measuring themselves on moving the needle in adoption now. I'd hate for such amazing progress to stall out in a niche.
Cloud resources are trending towards consumer technology adoption numbers rather than being reserved mostly for Enterprise. This is the most exciting thing in decades!
There is going to be a GPU/Accelerator shortage for the foreseeable future to run the most advanced models, Gemini 2.5 Pro is such a good example. It is probably the first model that many developers i've considered skeptics of extended agent use have started to saturate free token thresholds on.
Grok is honestly the same, but the lack of an API is suggestive of the massive demand wall they face.
1. Model "performance" judged by proxy metrics of intelligence have improved significantly over the past two years.
2. These capabilities are yet to be stitched together in the most appropriate manner for the cybersecurity scenarios the author is talking about.
In my experience, the best usage of Transformer models has come from a deep integration into an appropriate workflow. They do not (yet) replace the new exploration part of a workflow, but they are very scarily performant at following mid level reasoning assertions in a massively parallelized manner.
The question you should be asking yourself is if you can break down your task into however many small chunks that are constrained by feasiility in time to process , chunk those up into appropriate buckets or even better, place them in-order as though you were doing those steps with your expertise - an extension of self. Here's how the two approaches differ:
"Find vulnerabilities in this code" -> This will saturate across all models because the intent behind this mission is vast and loosely defined, while the outcome is expected to be narrow.
"
(a)This piece of code should be doing x, what areas is it affecting, lets draw up a perimeter
(b) Here is the dependency graph of things upstream and downstream of x, lets spawn a collection of thinking chains to evaluate each one for risk based on the most recent change . . .
(b[n]) Where is this likely to fail
(c) (Next step that a pentester/cybersecurity researcher would take)
"
This has been trial and error in my experience but it has worked great in domains such as financial trading and decision support where experts in the field help sketch out the general framework of the process where reasoning support is needed and constantly iterate as though it is an extension of their selves.
The DNI has officially declared that no classified information was in the signal thread. The whole thread is public now.
The information about a sensitive investigation, clearly not meant for anyone outside the White House seems to be available freely to the Guardian. How come? Where is the line if any?
The whole point of the Senate testimony from the DNI was that Signal was an approved application that comes pre installed on Government issued devices - and yes indeed, for secure comms.
Even Teams flags external participants to a chat. How was a phone number not known to be within the government perimeter allowed to be added with no alarm to a chat thread in an app pre installed and approved by the agency ?
There are more questions than answers here and its clearly suspicious to say the least that a prominent threat vector such as a mistaken phone number could go unnoticed and not trip a single flag. We're not talking about compromised sim cards or anything, a simple fat finger could expose a secure messaging app thread to an external participant and this is approved by the department for years? How many "Mistakes" over the years have gone unreported ?
Waltz or anyone on that thread isn't responsible for IT, so who ultimately didn't secure this vector?
So the Guardian getting access to internal White House deliberations and a FORENSIC INVESTIGATION, is not cause for concern but everything that people you don’t like , is, apparently.
This is precisely why the political divide is impossible to bridge. Everything I said indicates seriousness about classified information or even simply unauthorized access to information in such a cavalier manner that it’s published in the Guardian. Somehow, calling it out is more problematic than achieving a political end.
How does the Guardian have such intimate details of a forensic investigation at the White House level and in the same breath claim that unauthorized access to non public information is a threat to national security ?
It makes no sense for the media on one side of the political spectrum to claim the right to unfettered access to secrets.
Being able to fly international in Business Class for annual vacations should you choose to is a choice many would consider 100x better than $10 University food.
A big issue at the core of this phenomenon is the Fed funds rate and access to capital.
It has never been clear to me how having a single universal funds rate for the entire economy would prevent this very scenario of diverging fortunes between those with large capital reserves and those without. In the name of fighting inflation, small businesses now struggle for access to cheap credit and have been under stressful conditions for years now while richer American entities continue to bid up prices of luxury items and average home prices. The reality on the ground is far more dire than what the headline metrics show as we see an entire generation push up average duration of long term credit like car loans or home loans and thus set themselves back from otherwise needed population milestones by a decade or more right now.
The old cliche about "The first Million dollars is the hardest, but the next Million is an eventuality" comes from this very phenomenon we're seeing.
There has to be a better way for the Fed to have a fine grained approach at applying friction to money supply and credit rather than hitting everything in the economy at once.
I'm not sure why there is so much poor reporting on accelerator demand recently, it seems there are a lot of people looking to sell a message that isn't grounded in reality.