TWIT 802: The Year in Review

Beep boop - this is a robot. A new show has been posted to TWiT…

What are your thoughts about today’s show? We’d love to hear from you!

2 Likes

Fun show!

Leo accidentally saying “The” Ukraine made me laugh. I game with a Ukrainian buddy and he always gets all spun up when I say “The” Ukraine :grin:

1 Like

Excellent group of guests! @thurrott, @JeffJarvis, and Steve Gibson deserve to have their own show together!

And listening to the show put me in the mind that we really do have it good here on TWIT with the stellar quality of shows and hosts. So thank you @Leo for continuing to do what you do with TWIT and the shows. When I think about the few things I may want to complain about - all they really are is just personal preferences. What you have created is something special and I cannot thank you enough for all of the hard work and the team at TWIT do.

Have a Merry Christmas and a most excellent New Year!

6 Likes

Leo and the crew don’t notice the silencing of conservative voices on social media because they don’t follow those people. What happens when the fact-checkers are wrong? If you have a powerful enough voice and deep pockets you can sue them like Candace Owens is doing. Glenn Beck multiple times has had posts fact-checked incorrectly and after having his lawyers fight Facebook for a week they will admit they are wrong and say they will take the strike away. How much money did the New York Post lose when social media decided to kill the Hunter Biden story off their sites? Only now that the election is over it’s all coming out that it was all true. Notice they didn’t fact check the Russian Collusion farse.

If you ask me Section 230 should only apply to social media if they allow all legal speech without the fact-checkers getting in the way.

2 Likes

Great show as usual @Leo. Thank you for all the years of informative and entertaining content. It’s truly a balance with a great host and great guests. I remember there used to more frequent appearances of the other TWIT hosts on the big show. Did that not work out well? I really enjoyed hearing Jeff, Paul and Steve talk about things outside of their bailiwick. I, for one, would love to see more of it. Jeff is really coming along as a true co-host.

Great job as well @JasonHowell and everyone else for all the work you do.

Happy Holidays and a Happy New Year to all in the TWIT community.

Lesson of 2020: Ignore the trolls folks.

5 Likes

Fantastic point ya made…

I find perverse that the argumentation in Jeff’s referencing at ~01:16:00 the court ruling that blocked a TikTok ban framing it as a 1st Amendment case for interfering with the means by which individuals have their voice to engage the public IMO should apply all the more forcibly to domestic ISPs, yet their own, paying customers are considered to have no standing to sue for equitable terms of internet access that affect far more than the ability to engage the public sphere with speech but instead impact vital basic services like paying their own ISP and other utility bills, banking & ecommerce, etc., made even more critical under covid for access to education! The US obsession with entrepreneurial capitalism is an endometriosis seeing a foreign potential Macebook as justifying greater ostensible concern for US citizens’ Constitutional rights than any home-grown “threat” to extractive business models’ exploitation having far worse impacts on so many more citizens than even know what TikTok is!

1 Like

Also, this is, allegedly, being banned because it is giving information to the Chinese government, as opposed to the US government. I think it is bunkum.

Banning TikTok. however, is not a 1st Amendment issue, there are plenty of ways to have your voice heard, whether it be on a street corner, a theatre, Facebook or your own website…

1 Like

I couldn’t agree more (not that you’re trying to say I disagreed with that, but for clarity I appreciate your statement); I just wanted to point out what I feel to be the incoherence and contradiction in the argumentation of US regulators and politicians. Furthermore, it amazes me that the fear of deplatforming in the case of social media companies seems nowhere to be found in debates advocating against civil rights impingement through ISPs’ anti-competitive and predatory business practices, because the government is who would regulate it (all the while China is supposedly the enemy only the government can save us from, apparently? Actual encryption, actual privacy, actually social media rather than a free-for-all casino of attention-seeking on a private, profiteering 3rd-party’s “platform” would render moot almost all of the fears around TikTok vis a vis China, but advertizers would have a heart-attack and no one could leverage the presumption that everyone sees it to provide their B.S. performative oxygen).

3 Likes

In the discussion of “self-driving” at roughly the 2 hour mark, I have 2 things I’d like to note:

Steve Gibson’s practical approach perfectly pillories the absurd framing of the entire subject as being a subset of “AI”: yes, there needs to be intensive processing of the environment, but to cut it whole-cloth and task it with a human driver’s capabilities is not just utterly absurd as the panel richly chortles in absolutely justified scorn, but also needlessly hamstrings perfectly achievable imprevements to transportation and offloading of obscene amounts of wasted drive-time that could much more reasonably be had than if we set the floor at an “intelligence” general enough to synthesize the entire driving endeavor on-the-fly. It’s a bit of a vague point given the current nebulousness of the field, but for that is in fact all the more vital in its critique, IMO.

Jeff Hawkins’ Hierarchical Temporal Memory in Numenta brought to mind FPGAs and I’m curious whether or not Apple’s forthcoming generations of its own hardware will switch to such architecture in light of NVIDIA’s recent purchase of ARM IP.

In a bit of a sloppy triangulation of the 2 above musings, a further: a less fraught tac than Gibson’s tentative suggestions about remaking roads to provide guardrails/breadcrumbs for “autonomous vehicles” (or whatever to term them) could be a sufficiently fine-grained, curated representation such as already being undertaken by maps initiatives from Google and Apple, provided sufficient access to and negotiation with that model and the inevitably endless ongoing conditions about which vehicle and cloud should keep current (e.g. traffic, weather). No need to tear up roads or other laborious, red-tape laden disruption if instead it could all be done in software, eh? I realize the idea of “self-driving” is to synthesize it all in the moment, but again that’s an impossibly tall order that needlessly leaves so much on the table.

For example: a stop-sign is defaced and unrecognizable to “AI” “computer vision” “algorithms”. Who cares, if there’s supposed to be a stop-sign there and the car’s routines respect it even if it’s been flattened completely, stolen and removed altogether, or what have you (and it can react to the fact it’s missing with extra caution since other drivers may be liable to skip it, themselves, etc.: the fact that it’s missing could inform caution even if an exact diagnosis and assessment of the situation may be beyond its capabilities)?

The irony is, the technology for self-driving vehicles has been there for a couple of decades… What is still the biggest problem and the stupidest part of the equation, getting the technology to a level where it can interact with non-self-driving vehicles is a much harder problem.

So, we need to make the cars much more intelligent than needed for the first 20 - 30 years, then they can be dumbed down again once the normal vehicles disappear from the road.

If we turned the city-centres into autonomous vehicle only zones, those vehicles wouldn’t have to be half as clever as what we currently have. You “just” need to change the roads to allow the vehicles to recognize where they are and they need to communicate with each other. No more traffic lights, no more traffic jams, no more accidents (unless a vehicle fails, then it should fail “stopped” and other vehicles should route around it).

You “only” have to take into account pedestrians and people hacking the street furniture to confuse the vehicles. Dealing with irrational drivers trying to cut up other road users to get a car length in front disappears, drunk drivers disappear, idiots behind the wheel disappear, big egos behind the wheel disappear, oh, sorry, repeating myself there…

I think this is a Faustian bargain in effectively forfeiting freedom of movement to a vast degree. Technology’s role, IMO, must be circumscribed to accommodation within, and not mastery over (through regulation, incompetence creating untenable danger deemed preferrable over continuing to allow human drivers, or otherwise), transportation. Another important aspect to consider is the accommodations beyond cars, e.g. for e-bikes, powered skates/skateboards, and other options which may be within fiscal and logistical reach for the less enfranchised, not reliant on a 3rd-party broker beyond initial sale and some maintenance.

1 Like

They already do this sort of thing – use mapping and other “cloud” data to navigate – but that cannot ever obviate the need for direct perception of the local environment. Mapping data can be outdated; a path could be open that the data is saying isn’t, or worse, a path could be closed that the data is saying is open. There could be things – or people – in the road. And of course the system has to track all the other cars.

Trains have literal guardrails – or rails anyway – but that doesn’t obviate the need for an operator, human or machine.

While the technology for in-the-moment synthesis is cooler, and would bring the “intelligence” of self-driving cars up to the level of human drivers, I don’t think that’s actually an inherently important factor for the commercialization of self-driving cars at all. Just as it would be cool if we could make a self-driving car that only depended on two cameras and two mics on a round rotatable mount for perception, instead of the array of sensors they currently have. Yes, cool, but it’s not necessary to reach that level prior to commercialization.

If massive amounts of data can be crunched and processed and made available for the car to then use to “stupidly” drive in the moment, that would be sufficient. But as I said above, there’s still a lot of on-the-ground computation that needs to be done.

I heartily agree. It’s a pretty abstract discussion, still, at this point, given how primitive what’s being offered still is. We’re debating it only because of its being so over-sold, I’m sure. I dealt with the case of data that should be there being instead absent, but yours is the more pressing case of data that’s actively mis-matched. I’m also engaging the subject on the terms of the products being discussed in the show’s segment that rely on computer “vision” of the moment (do they even use LIDAR?). It’s a very modest, largely rhetorical tweak to Gibson’s point I seek to make, nothing more😅

In major cities, at least around here, bikes and scooters etc. are already segregated from general road traffic. City centres are becoming more and more pedestrian only anyway. You have to park on the outer ring, or further out, and walk, take a bus, tram or train into the centre. Commercial deliveries are usually limited to being before 07:30 in the morning.

Between towns, I think the technology just isn’t currently there. But removing human drivers from the chaos that is city life would make a huge difference to the efficiency of cities. Do you really want to sit for an hour behind the wheel in a traffic jam, instead of 10 minutes in an automated vehicle, whether that be a bus or a taxi, just because you can drive yourself? Many people already use buses, taxis, Ubers etc., so they aren’t driving themselves anyway.

The US suffers from the worst of both worlds: no superpedestrian infrastructure (for e-bikes, etc.) and painfully few city centers dense enough for pedestrian-only to work nearly so well (without a swarm of secondary transport such as scooters or powered skates) as in European cities that accumulated their design at human, rather than automobile, scale. The danger I see in your formulation is that the vision of automated transport will subsume the important details of their trade-off’s for individuals for whom that represents a potentially high barrier in costs to privacy of a 3rd-party vehicle/service (and/or money/price), and where accommodations for alternatives Europe’s been smart enough to allocate for properly are left whatever crumbs remain of space and safety.

2 Likes

There is already a prototype service running around the grounds of the Charité hospital in Berlin. A bus runs around on a pre-configured route, it stops if people walk out in front of it and anybody can board or leave at any time.

Some cities are moving to free buses within the city limits, whilst trying to encourage drivers to leave their vehicles at home or at park and ride spaces outside the cities. Park and ride has been big here for decades. You drive up the Autobahn until you approach the city, then there are P+R signs around the area, near the railway stations and you take a train or bus into the city itself.

You only need to actually drive into the city if you have ordered something bulky that can’t be carried (E.g. new TV). And the, you usually have pick-up points behind the shops, as the fronts are in pedestrian zones.

It is second nature here, to drive to the edge of town and park up, then walk or take public transport into the centre.

I can’t remember the last time I took a taxi, I think, when I was still living alone and was ill and needed to get to the doctor’s. So that would have been 2009, probably.

I used to commute to a job in the local city. I’d walk the 20 minutes to the station, here, then 20 minutes to the city and then walk through the city to the offices, another 15 minutes. Driving at rush hour would have taken at least 40 minutes and the cost of parking in the city would have been more than the train ticket.

3 Likes

I worry that the rise of autonomous vehicles in the United States will only further entrench the terrible auto-centric mindset that has guided zoning, funding, and policy decisions the past half century and led us away from denser, incremental development that can more flexibly accommodate not only more people but more possibilities and economic opportunities.
@big_D O I wish we had more communities built around public transportation, public places, and the public in general.

4 Likes

The 20th-Century thinking of the US would charge micropayments every time you slowed the bus down even as a pedestrian, and much more for touching it, especially as a passenger, then force the bus service to run exclusively off the revenue generated, because entrepreneurial private capitalism🤮 Europeans have a clue about civic infrastructure and the topic of state benefits isn’t radioactive.

This is why I pushed back so hard against Alex Lindsay’s glib dismissal of existing US cities in MBW #686: in the seeds of the dismissal are the toxins of myopia that led cars to so dominate in the past century, and there are surely at least as many traps to be developed in “smart cities” that the same impatience with humility cannot but ensure take root and strangle the lives of residents, especially those not centered in the socioeconomic “design doc”.

3 Likes

I think Steve Gibson had the best line: “I’ve got enough G’s as it is!” So true!

1 Like