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.
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.
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.
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.
I would say they're winning with Waymo: I took a fully autonomous taxi ride in the backseat in SF, and it just worked. No other company can currently do that, despite their promises and hype.
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?
As with everything related to Tesla FSD/Autopilot, I'll believe it when I see it. They have not earned the benefit of the doubt. Waymo works as a robotaxi today, Tesla doesn't.
I'll grant you Chinese developments; I'm not across what's happening there, but I wouldn't be surprised if it was on par, yes.
My bet is that they can reduce the cost of their working solution more reliably and safely than Tesla can get their solution working at scale.
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)
Winning going on what, 10 15 years now? Surely at some point they must start scaling?
At this point all I can imagine is that every year they run the numbers and arrive at "yup, still makes no sense whatsoever". And so its eternally doomed to tech demo status.
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.
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.
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.
No, you are not differentiating between information that is classified and information that is merely non-public.
This is a difference in law, there is a difference in duty of care (although even then, the Trump administration is responsible in both cases), and an enormous in impact (mission failure vs Trump embarrassment).
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?
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.
Sure. I didn’t make my point clear.
100x expense doesn’t mean 100x utility. There are a lot of “consultants” in HCOL cities grifting from high earners.
I had a baby proofer try to charge me 4 grand for a handful of outlets and corner pads.
yah, rich people are extremely stupid in ways that doesn't really hurt them because it's a small portion of their income. How jimbo might do something stupid and spend 10% of his disposable income on scratchers trying to get rich, rich people do the same, it's just 10 % of their income is a large number and they are usually chasing convenience or experience rather than riches.
