Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bobsil1
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
301.
▲
by
bobsil1
9y ago
Waymo is more server-side, requires hi-res maps. Tesla's approach should be more adaptable long-term. Their AP 2 is actually In-House AP 1.0, of course it does less than MobileEye v 15.0 or whatever.
302.
▲
by
bobsil1
9y ago
They're bulk uploads over home WiFi, and Tesla hired plenty of ML names like Andrej Karpathy. Having said that, AGI is the bottleneck for full self-driving.
303.
▲
by
bobsil1
9y ago
Sensors help, but the real bottleneck for edge cases is artificial general intelligence.
304.
▲
by
bobsil1
9y ago
The edge cases are endless (wheelchair woman chasing a chicken), and without a semantic world model the car will have no idea what to do. Low-level object avoidance architecture isn't enough.
305.
▲
by
bobsil1
9y ago
The edge cases are unclear: construction zone, heavy equipment, school kids, another car poised to enter the road.
306.
▲
by
bobsil1
9y ago
Default PIN has been 6 digits since iOS 10, and you can make it a long passphrase instead.
307.
▲
by
bobsil1
9y ago
>the mass market does not really care whether a car is gas or electric Which is why gov'ts need to ban ICE sooner rather than later, to save summers and stave off species suicide.
308.
▲
by
bobsil1
9y ago
LUWTV, lube up and wait to vest
309.
▲
by
bobsil1
9y ago
More like the Winter of Artificial General Intelligence
310.
▲
by
bobsil1
9y ago
Face ID is going to screw this all up. Cops won't even need your cooperation to unlock.
311.
▲
by
bobsil1
9y ago
MS was always in bed with the feds, e.g. Skype and crypto backdoors.
312.
▲
by
bobsil1
9y ago
>the "last 1%" of self-driving software may be equivalent to solving AGI. Exactly right. You need a semantic world model, reasoning and prediction. AGI bottlenecks self-driving, assistants, text understanding…
313.
▲
Cruise's 3rd generation self-driving car
(medium.com)
201 points
by
bobsil1
9y ago
|
214 comments
314.
▲
by
bobsil1
9y ago
Where she got the idea from: https://2.bp.blogspot.com/-_WjsocrjdnA/V5rP7GnI8pI/AAAAAAAAA...
315.
▲
by
bobsil1
9y ago
Big spender. Using my Vic-20
316.
▲
by
bobsil1
9y ago
That was the paper I was thinking of (thanks), but maybe the figure is from an IBM True North paper. I vaguely recall 10^-5 for brain efficiency, 10^-4 for True North vs. von Neumann.
317.
▲
by
bobsil1
9y ago
Another Slack clone that's not end-to-end encrypted. Why even bother?
318.
▲
by
bobsil1
9y ago
Human brain uses 10^5 less power than von Neumann architecture, according to Caltech paper from ‘90s.
319.
▲
by
bobsil1
9y ago
Naysayers are a dime a dozen. What's valuable: someone who can see how to make it happen.
320.
▲
by
bobsil1
9y ago
Hand-engineering and hand-assembling evolved components will fail, way too complex. We struggle with making bug-free word processors. This will only be solved by fully evolving AGI, or some other form of automated search of high-dimensional
321.
▲
by
bobsil1
9y ago
Why reverse it when you can evolve it?
322.
▲
by
bobsil1
9y ago
Salary slashes are inevitable because healthcare costs are growing like metastatic cancer. The twin dragons are health and milsec.
323.
▲
by
bobsil1
9y ago
After a long history of fatalities and fixes.
324.
▲
by
bobsil1
9y ago
My novelist friend behind this writes 700-900pp books and complains he has to switch between Word, Project and Excel to keep everything straight. They're hoping it's useful for complex universes, fan fic and investigative journali
325.
▲
by
bobsil1
9y ago
Depends on the task. It's made to automate continuity for long novels, complex fiction/fan fic universes and complex investigative journalism.
326.
▲
by
bobsil1
9y ago
Totally agree. Check out Granthika, a semantic novel editor with automatic error checking for timeline and character errors: http://granthika.co In progress, pre-beta, by a friend and novelist.
327.
▲
Granthika – semantic novel editor with auto timeline/character error checks
(granthika.co)
36 points
by
bobsil1
9y ago
|
3 comments
328.
▲
by
bobsil1
9y ago
Desirable EVs which define the category.
329.
▲
by
bobsil1
9y ago
Until one shot and transfer learning work.
330.
▲
by
bobsil1
9y ago
Hard tech? Righetti quantum computing, IBM Research neuromorphic chips, Geometric Intelligence one-shot ML…
More ›