The consistent reports I'm getting through friends who have spoken to recruiters is just that standards have increased dramatically. Firms accepting only a single candidate CV from a recruiter instead of 5 or more, and the recruiter downstream working their ass off to ensure it is absolutely the best possible CV, because if it's not then they don't get paid.
SpaceX is a different beast with extremely high friction to enter its market, a massive technology lead, and well developed preferential high level relationships with just about every country worth worrying about.
OpenAI/Anthropic meanwhile feel a bit like they're hoping to sell iPhones in a market about to be flooded by $20 flip phones, with almost no channel of their own to do it. And for whatever mad reason OpenAI are now signalling they will attempt to compete on price with flip phones despite their cost of labour, energy, and just about everything else being far higher
It's asking a lot to trust they can or will maintain this new pricing. In any case it's exciting to think this might lead to further price cuts in the highly competent and competitive Chinese clones. I'm still using ChatGPT for interactive queries, but at this point pretty much only because of its familiar UI
It reminds me a bit of what Ansible got right: user communication. The underlying tech may have existed for a long time, but the genius is presenting it to a regular developer in a way that reads "yes, even you can understand ML, just using a little JSON". The contribution of that should not be understated, as has been clearly evident recently.
Everything is always better with more YAML, are you perhaps new to this industry?
Next we also need an instruction style guide and CoC. It's important to treat your agents with respect. I almost forgot, the YAML template meta-language to YAML the YAML. Then we will need a foundation employing 12 FTEs to maintain it all and of course to run the certification process. You are certified, right? Statistics show a 10x increased chance of an agent going rogue and hacking competitors if it has been mistreated or been run in an unvalidated sandbox. It goes without saying the sandbox certification process is separate and must be repeated yearly by a trusted third party auditing company.
You know it's underdeveloped because everything is described as "basic" or "simple". There's no such thing in any complex design, yet somehow for all his purported seniority, he keeps pumping out designs using the same description.
Where "basic" follows a highly nuanced alternate definition, where if you step out of line without realising it, you lose 3 days debugging a complex undocumented interaction between 4 enterprise crapware daemons only to find out the answer is "you're wrong and won't fix" buried in a 6 year old bug report, when all you did was plug in a USB mouse or try to set the default audio output device, or something else.
Total and utter propaganda, it's underdeveloped steaming garbage
Utterly moronic. We support the freedom of the Iranian people by forcing them to install a local government root CA in every browser. I mean at this rate they won't even have to buy their monitoring tech from China any more, just an old PC and a late 90s tarball of Squid
I imagine they're doing this due to capacity issues or somesuch. They can always relaunch Pro later, meanwhile a little ricered benchmaxxing of their existing flash model provides a temporary cover story. They certainly aren't silly enough to think this won't impact existing Pro users </paranoia>
I think the most enjoyable part of this is to have written it in the characteristic long-form Zitron wall of text style while still managing to pack almost the whole text with meaning, which is entirely the opposite of Zitron style.
Curious what MCP setup you use? I'm not sure which one I have wired up, but I have to restart Ghidra every time I change file. I think it's either LaurieWired's original or a fork of it
You should find an excuse to offer 3D printed extruded pelicans from various models as awards for something. I have no idea for what, but the idea captivates and I'd love to win one somehow. They'd be collector's items in a few decades
DS4 Flash absolutely kicks ass for reverse engineering and bug hunting. Almost no point in considering paying for a bigger model, although it's possible the stuff I've fed it (wide variety of older DOS/Windows stuff and device firmwares) might be easier targets.
Using codex every day, in spite of which, I hope some day providers will just start naming their offerings small/medium/large, a bit like we eventually started doing in software testing. Trying to remember what Sol is or why it's better than the other thing is more cognitive effort than I can muster at this point. And that's a sure sign of commoditisation in itself
Looks like an awesome release, but the smell of AI from that post is horrid.
Here is a wild idea: is it really so hard to edit out sentences structured and punctuated like this - it's so painfully obvious and distracts from the content. The effect is real.
There is plenty of historical data, $100k is a steal for as long as situations like Iran persist. And now he is incentivised to ensure they do persist. Our fearless leader, ladies and gentlemen.
They were too small for this to be a meaningfully sized purchase for AMD, there's real risk they get sucked into a team that ultimately delivers sqat, not to mention the chances of anything being delivered in an even remotely consumer-priced bracket are definitely out the window
If CADA passes in its current state I doubt we'll be banning Chinese models any time soon. I think most people misunderstand the extent of the permanent damage caused by Trump
Never worked with a security department that volunteered pretty much anything. Anthropic and OpenAI's meanwhile, they have simultaneous timed releases in the BBC and FT and presumably a dedicated PR manager.
Whoa there Amodei, are you saying your model hacked into a super secure environment too? How alarming and incredibly well timed.