Same way that I would trust your review to be accurate. Because the reviewer has built a reputation for correctness.
Its not Claude doing the review. Its a human doing the review, but using Claude to do the reading. Its still on the human to ask the right questions to Claude.
This is a blog on the root cause. MCAS would be an intermediate mechanism in making you feel sick, but something must have triggered the MCAS. Thats the autoimmune response.
Last week I built Grove as the successor to my IDE. Like many, I suspect, I was struggling to coordinate multiple agents per repo recently. Cursor, Zed, Superset, Conductor, all are now following a layout centered around worktrees, but I felt like they were missing the point. I just need some organizer for my shells and a diff viewer. I no longer want to deal with a debugger, or an lsp, or code navigation.
Give it a whirl if this resonates with you! Grove is rough around the edges and was designed with my workflow in mind, but I welcome contributions to eg supporting more VCS, another package manager, or be more friendly to whatever agent you like!
Built in Rust, because why not? An Electron stack just sounds like a hassle to maintain in this Brave New World of ours.
I want this, but I do not have the experience with radio signals to build this myself without more guidance. Is there a DIY proof of concept I could lean on? How much more challenging will this be if you are in an area with overlapping FM signals from 2 transmitters sending the same signal?
I feel like your comment is in itself a great analogy for the "beware of using LLMs in human communication" argument. LLMs are in the end statistical models that regress to the mean, so they by design flatten out our communication, much like a reductionist summary does. I care about the nuance that we lose when communicating through "LLM filters", but others dont apparently.
That makes for a tough discussion unfortunately. I see a lot of value lost by having LLMs in email clients, and I dont observe the benefit; LLMs are a net time sink because I have to rewrite its output myself anyway. Proponents seem to not see any value loss, and they do observe an efficiency gain.
I am curious to see how the free market will value LLM communication. Will the lower quality, higher quantity be a net positive for job seekers sending applications or sales teams nursing leads? The way I see it either we end up in a world where eg job matching is almost completely automated, or we find an effective enough AI spam filter and we will be effectively back to square one. I hope it will be the latter, because agents negotiating job positions is bound to create more inequality, with all jobs getting filled by applicants hiring the most expensive agent.
Either way, so much compute and human capital will go wasted.
I am stumped. Am I misreading, or are the folks at Google deliberately confounding two interpretations of "world model"? Dont get me wrong, this is really cool, and it will undoubtedly have its use. But what I am seeing is an LLM that can generate textures to be fed into a human-coded 3d engine (the "world model" that is demonstrated), and I fail to see how that brings us closer to AGI. For AGI we need "world models" as in "belief systems". The AI model must be able to reason about (learned) dynamics, which I dont see reflected in the text or video.
I see at least 2 axes here:
* Should access to a tool be restricted of it is used for malice
* Is a company complicit if its automated service is being used for malice
For 1, crowbars are generally available but knives and guns are heavily regulated in the vast majority of the world, even though both are used for murder as well as legitimate applications.
For 2, things get even more complicated. Eg if my router is hacked and participates in a botnet I am generally not liable, but if I rent out my house and the tenant turns it into a weed farm i am liable.
Liability is placed where it minimises perceived societal cost. Emphasis on perceived.
What is worse for society, limiting information access to millions of people or allowing csam, harrassment and shaming?
It is not meant to save the doctors face. The very definition of FND is "doctors dont know what is wrong, but they acknowledge that your symptoms are real".
The point of giving it a name is in the second part. Its about explicitly acknowledging the limitations of medicine
I understand your frustration about the "I'm from Europe" comments, but you have to appreciate that those comments come from a similar frustration.
The universally shared image of the US is that they loudmouth how great they are compared to the rest of the world. So when news comes out on how the US fails so dramatically on very basic public/social services, it leaves people from outside the US stumped.
If you're asking how widely available SIMD is, it has been common in consumer hardware for 2 decades. To perform SIMD instructions manually you will need a compiled language that supports it, like Rust or C. But the compiler can actually implement it for you as an optimization.
What I'm getting from the article: it is a tool that advises affiliate marketeers on which affiliate links work, and which ones are wasted effort. I'm not 100% sure as well though.
I dont know what I expected. It's a 17 year old earning 600 bucks per month on an affiliate marketing tool.
Very cool! Although I would love to be proven wrong, I am still suspicious that this is actually a code sample from an existing game it picked up somewhere during training. Seeing how chatgpt struggles with basic logic I would be surprised if it can actually generate a new and workable game of logic.
I'm missing one very important cost on the contractor side: unemployment insurance. I don't know about German law but in the Netherlands all the numbers are comparable, except that your employer is required to insure you against loss of income. This insurance runs up to 1/3 of your insured income, which would put your taxable income at about 87.000. Of course you don't _have_ to take this insurance but if you want an honest comparison it should be included on both sides.
Finding the right architecture, or more in general the right model, is very much still the main problem.
You should be careful with the meaning you ascribe to the word 'universal'. The list of universal approximators is massive, and the sub-list of universal approximators that can be trained with OLS is still substantial. Still these models can differ significantly:
- How efficient are they (in #parameters required for a certain error) for specific tasks? There is a known 'maximum efficiency' for general tasks, but in high dimensions this efficiency is terrible, such that many models will fail terribly on high-dimensional data. Hence, you should pick a model that is exceptionally good for a specific task, although it might be less efficient for other tasks.
- How well can the model cope with noise? If your dependent variable is severely distorted (think financial data) then you need a model that can balance between interpolating datapoints and averaging out the noise.
Just to name my two favorite properties. The first one is _kind of_ related to learnability, since an inefficient model is often pretty much impossible to learn.
No monetization of any kind other than a slightly hidden donate button.