Results for “direct instruction” 42 found
Lessons from the COVID War
In preparation for a National Covid Commission a group of scholars directed by Philip Zelikow (director of the 9/11 Commission) began interviewing people and organizing task forces (I was an interviewee). The Covid Commission didn’t happen, a fact that illustrates part of the problem:
The policy agenda of both major American political parties appear mostly undisturbed by this pandemic. There is no momentum to fix the system….The Covid war revealed a collective national incompetence in governance….One common denominator stands out to us that spans the political spectrum. Leaders have drifted into treating this pandemic as if it were an unavoidable national catastrophe.
The results of this early investigation, however, are summarized in Lessons from the COVID WAR. Overall, a good book, not as pointed or data driven as I might have liked (see my talk for a more pointed overview), but I am in large agreement with the conclusions and it does contain some clarifying tidbits such as this one on the Obama playbook.
Innumerable speeches, books, and articles have stated that the Obama administration gave the incoming Trump administration a “playbook” on how to confront a pandemic and that this playbook was ignored. The Obama administration did indeed prepare and leave behind the “Playbook for Early Response to High-Consequence Emerging Infectious Disease Threats and Biological Incidents.”
But this playbook did not actually diagram any plays. There was no “how.” It did not explain what to do…when it came to the job of how to contain a pandemic that was headed for the United States in January 2020, the playbook was a blank page.
I also appreciated that Lessons has some some unheralded success stories from the state and local level. You may recall Tyler and I blogging repeatedly in 2020 about the advantages of pooled tests. Eventually pooled testing was approved but I haven’t seen data on how widely pooling was adopted or the effective increase in testing capacity that was produced. Lessons, however, offers an anecdote:
In San Antonio, a local charitable foundation paired with a blood bank to create a central Covid PCR testing lab (antigen tests were not yet readily available) that could combine samples (pooling) for efficiency and cost reduction, but also determine which individual in a pool was positive. Importantly, results were available within about twelve hours. That meant results were available before the start of school the new day.
The program helped San Antonio get kids back into the schools.
More generally, it’s striking that US schools were closed for far longer than French, German or Italian schools. See data at right on the number of weeks that “schools were closed, or party closed, to in-person instruction because of the pandemic (from Feb. 2020-March 2022)”. (South Korea, it should be noted, had some of the most advanced online education systems in the world.)
One general point made in Lessons that I wholeheartedly agree with this is that the school closures and many of the other controversial aspects of the pandemic response such as the lockdowns and mask mandates “were really symptoms of the deep problem. Without a more surgical toolkit, only blunt instruments were left.” With better testing, biomedical surveillance of the virus and honest communication we could have done better with much less intrusive and costly policies.
Addendum: See my previous reviews of Gottlieb’s Uncontrolled Spread, Michael Lewis’s The Premonition, Slavitt’s Preventable and Abutaleb and Paletta’s Nightmare Scenario.
Addendum 2: A typo in Lessons had France closing schools for 2 weeks instead of 12 weeks. Corrected.
How on-line education affects grading
This paper examines the role of student facial attractiveness on academic outcomes under various forms of instruction, using data from engineering students in Sweden. When education is in-person, attractive students receive higher grades in non-quantitative subjects, in which teachers tend to interact more with students compared to quantitative courses. This finding holds both for males and females. When instruction moved online during the COVID-19 pandemic, the grades of attractive female students deteriorated in non-quantitative subjects. However, the beauty premium persisted for males, suggesting that discrimination is a salient factor in explaining the grade beauty premium for females only.
That is from a new Economic Letters piece by Adrian Mehic. Via tekl.
The cost of applying
- Rhodes Scholarships: 12,000 applications, 99.3% rejected
- National Science Foundation: 11,447 applications, 82.7% rejected
- National Institutes of Health: 55,038 applications, 79% rejected
- Nonprofit grants: too many to count, but one professional grant writer estimates 90% get rejected
These applications are hefty. Before you even start applying for an NSF grant, you should probably read the 79 pages of instructions. The NIH helpfully provides a 10-part instructional video series. Rhodes Scholarships require a mind-boggling eight letters of recommendation. I tell college seniors to expect fellowship applications to be a six-month part-time job.
All this applying doesn’t just burden applicants. Professors run themselves ragged writing recommendations. The NSF relies on volunteers to complete 240,000 reviews every year. Entire university offices exist just to manage the paperwork that grants generate; universities bill this back to funders in the form of “indirect costs,” which at Harvard go as high as 70% of incoming grant funding. Grant agencies seem not to realize that by making everything about their grants burdensome, they allow universities to spend much of the grant money managing the grant itself!
Here is more from Adam Mastroianni. Via Anecdotal.
If UFOs are alien beings, are they just doing mood affiliation in visiting us?
Robin Hanson has a long and very interesting blog post on that question. The point is not to argue that the UFOS are alien beings of some kind, but rather if they were which kinds of theories might help us understand them? Here is just part of Robin’s much longer take:
If the main block to believing in UFOs as aliens is a lack of a plausible enough social theory of aliens, then it seems a shame that almost no one who studies UFOs is a social science theorist. So as such a person, why don’t I step in and try to help? If we can find a more plausible social theory, we could become more willing to believe that UFOs are aliens…
Stylized fact #2: Aliens are rare and self-limited, and yet are here now.
Indirection – We can think of a number of plausible motives for rare limited aliens to make an exception to visit us. First, they may fear us as rivals, and so want to track us and stand ready to defend against us. Second, if their limitation policies are intentional, then they’d anticipate our possibly violating them, and so want to stand ready nearby to enforce their limitation policies on us.
In either of these two cases, aliens might want to show us their power, and even make explicit threats, to deter us from causing problems. And there’s the question of why they don’t just destroy us, instead of waiting around. Third, independent alien origins could be a rare valuable datapoint about far-more-capable aliens who they may fear eventually meeting. In this case they’d probably want to stay hidden longer.
My best bet is this. The vehicles would be “unmanned” drone probes, if only because the stresses of long trips through space would keep the actual alien beings close to home. So the relevant social science question is what kind of highly generalized software instructions you would give such drones. “Seek out major power sources, including nuclear, and seek out rapid flying objects, and then send information back home” would be one such set of instructions roughly compatible with the stylized facts on the ground (or in the air). Of course the information sent back to alien worlds will not be arriving for a very, very long time, so long that the concrete motives of the aliens may not be the major consideration. Collecting the information about other planets across some very long time frame might simply seem worthwhile, relative to the cheap cost of the drone probes. It reminds me a bit of that “put the DNA of all the species on the moon” project we have started, or those seed banks up in the Arctic. Why exactly did we do it? Why not I say!? And yet most humans do not even know those projects are going on.
A further generalized software instruction would be “if approached or confronted, run away fast.” Indeed that is what those flying vehicles seem to do.
The drone probes do not destroy us, because of Star Trek-like reasons: highly destructive species already have blown themselves up, leaving the relatively peaceful ones to send drones around. The drones probably are everywhere, in the galactic sense that is. Yet given the constraints imposed by the speed of light, it is difficult to do much with them that is very useful to the decision-makers that send (sent?) them out. So the relevant theory is one of how advanced civilizations allocate their surplus when there is a lot of discretion and not much in the way of within-lifetime costs and benefits to determine a very particular set of plans and goals. Not even for the grandkids.
In this hypothesis, of course, you have to be short immortality. And short usable wormholes.
By the way, don’t those photos of the drone probes make them look a bit like cheap crap? No tail fins, no “Close Encounters of the Third Kind” music signature, no 3-D holograms, just a superfast vehicle. Like something a second-rate alien non-profit picked up at the local Walmart and sent off into space en masse with solar-powered self-replication. Which is consistent with the view of them being a discretionary resource allocation stemming from projects with fairly fuzzy goals.
A problematic question for any theory is whether competing drone navies have come to visit us, and if so are they fighting over the spoils? Colluding? Hiding from each other? Or what? If aliens are afoot, why should it be only one group of them? That would seem strange, as in most things there are multitudes, at least speaking in Bayesian terms. Aren’t there at least both Klingon probes and Romulan probes, maybe Federation probes too.
Robin’s hypothesis, that they are elatively local panspermiacs, who feel some stake in us, appeals to me. Bayesian logic suggests in any case that the chance of us having resulted from panspermia is pretty high; there are lots of baby civilizations for each parent, so why deny you are probably a baby?
Perhaps our visitors are exercising some “mood affiliation” in wishing to visit and record us! They could be the parents, or perhaps another baby civilization.
Of course since the photos are of such poor quality, and since there is no corroborating evidence of any kind, these UFO sightings probably are not of alien creations, so all of this is pure fantasy anyway.
2021 assorted links
1.Why a longer dosing interval should be fine. And the UK case for first dose prioritisation.
2. “We find that [Chinese] police stations are more likely to be located within walking distance of foreign religious sites (churches) than other sites (temples), even after controlling for the estimated population within 1km of each site and a set of key site attributes.” Link here.
3. Some UK doctors will defy instructions on postponing the second shot (NYT).
The farmer-influencer and the economics of streaming
Though Mr. Gold sells poultry and eggs from his duck farm in Vermont’s northeast corner, most of what he produces as a farmer is, well, entertainment.
Mr. Gold, who is short and stocky, with the good-natured ease of a standup comedian, does his chores while carrying a digital camera in one hand and murmuring into a microphone.
Then, twice a week, like clockwork, he posts a short video on YouTube about his exploits as a neophyte farmer, often highlighting failures or pratfalls. Keeping a close eye on analytics, he has boosted his YouTube audiences high enough to provide a steady advertising revenue of around $2,500 to $4,000 a month, about eight times what he earns from selling farm products.
And this:
It’s a lot of work: Mr. Lumnah wakes up at 3:30 a.m. so he can edit the previous day’s footage in time to post new video at 6 a.m., which his 210,000 regular viewers, who are scattered as far as Cambodia and India, have come to expect. “People will say, it’s lunchtime here in Ukraine,” Mr. Lumnah said.
Others, like Justin Rhodes, a farmer in North Carolina, have parlayed a giant YouTube audience into a dues-paying membership enterprise — he has 2,000 fans who pay annual fees of up to $249 for private instruction and direct communication, via text message. “We don’t sell a single farm product,” Mr. Rhodes said. “Our farm product is education and entertainment.”
Some of them earn money through product endorsement deals, like Al Lumnah, who posts videos five days a week from his farm in Littleton, N.H.
Here is more from the NYT, via Steve Rossi.
Some doubts about medical ethics, and maybe that Russian vaccine is underrated
Most major questions in ethics are unsettled, though of course I have my own views, as do many other people. I take that unsettledness as a fairly fundamental truth, I have been studying these matters for decades, and I even have several published articles in the top-ranked journal Ethics.
Now, if you take a whole group of people, give them medical licenses, teach them all more or less the same thing in graduate school, but not much other philosophy, and call it “medical ethics“…you have not actually gone much further. Arguably you have retrogressed.
So when I hear people appeal to “medical ethics,” my intellectual warning bells go off. To be sure, often I agree with those people, if only because I think contemporary American institutions often are not very flexible or able to execute effectively on innovations. For instance, I didn’t think America could make a go at Robin Hanson’s variolation proposal, and so I opposed it. “Medical ethics” seems to give the same instruction, though with less of a concrete institutional argument.
Still, the Lieutenant Colombo in me is bothered. What about other nations? Should we ever wish that they serve themselves up as medical ethics-violating guinea pigs, for the greater global good?
Medical ethics usually says no, or tries to avoid grappling with that question too directly. But I wonder.
How about that Russian vaccine they will be trying in October?
To be clear, I won’t personally try it, and I don’t want the FDA to approve it for use in the United States. But am I rooting for the Russians to try it this fall? You betcha. (Am I sure that is the correct ethical view? No! But I know the critics should not be sure either.) I am happy to revise my views as further information comes in, but I see a good chance that the attempt improves expected global welfare, and I think that is very often (but not always) a standard with strong and indeed decisive relevance. And all the new results on cross-immunities imply that some pretty simple vaccines can have at least partial effectiveness.
Why exactly is “medical ethics” so sure this Russian vaccine is wrong other than that it violates “medical ethics”? All relevant scenarios involve risk to millions of innocents, and I have not heard that Russians will be forced to take the vaccine. The global benefits could be considerable, and I do note that the Russian vaccine scenario is the one that potentially spends down the reputational capital of various medical establishments.
Trying a not yet fully tested vaccine still seems wrong to many medical ethicists, even if the volunteers are compensated so they are better off in ex ante terms, as in some versions of Human Challenge Trials, an idea that (seemingly) has been elevated from “violating medical ethics” to a mere “problematic.” Medical ethics claims priority over the ex ante Pareto principle, but I say we are back to the unsettled ethics questions on that one, but if anything with the truth leaning against medical ethics.
I find it especially strange when “medical ethics” is cited — often without further argumentation or explanation — on Twitter and other forms of social media as a kind of moral authority. It then seems especially glaringly obvious that the moral consensus was never there in the first place, and that there is a gross and indeed now embarrassing unawareness of that underlying social fact. It feels like citing Kant to the raccoon trying to claw through your roof.
I think medical ethics would not like this critique of medical ethics. Yet I will be watching the Russian vaccine experiment closely.
Addendum: There is also biomedical ethics, but that would require a blog post of its own. It is much more closely integrated with standard ethical philosophy, though it does not resolve any of the fundamental philosophical uncertainties.
GPT-3, etc.
Here is an email from a reader, I do not really have an opinion of my own yet. Please note I will not indent any further:
“I wanted to draw your attention to something. Are you familiar with “AI Dungeon,” text-based RPG “open world” game running on GPT-2 / GPT-3? Here’s the author’s discussion on medium, or you can play the GPT-2 version for free to get a sense of it directly.
But what I really want to draw your attention to is players who are using custom prompts to open up dialogs with GPT-3 about non-game things.
This result is particularly impressive: https://www.reddit.com/r/slatestarcodex/comments/hrx2id/a_collection_of_amazing_things_gpt3_has_done/fy7i7im/
( Following that string of posts is a little hampered by reddit’s format; here are the posts in order: part 1, part 2, part 3)
If the author is to be believed, they’ve had GPT-3 / “Dragon”:
1. write code
2. act as a pharmacology tutor
3. write poetry
4. translate english, french, chinese (the instruction to “balance the intent of the author with artistic liberty” is particularly interesting)
It’s hard to excerpt, I’d recommend reading the whole thing if you have time.
Here’s another user’s eloquent conversation about the experience of being an AI, using a similar mechanism (screencap images of the convo, part 1 and part 2 ), with a sample prompt if you want to converse with GPT-3 yourself via AI Dungeon.
I am increasingly convinced that Scott Alexander was right that NLP and human language might boostrap a general intelligence. A rough criteria for AGI might be something like (i) pass the Turing test, and (ii) solve general problems; the GPT-3-AI-Dungeon examples above appear to accomplish preliminary versions of both.
GPT was published in June 2018, GPT-2 in February 2019, GPT-3 in May 2020.
As best I can tell GPT -> GPT2 was ~10x increase in parameters over ~8 months, and GPT2 -> GPT3 was ~100x increase of parameters over ~14 months. Any number of naive projections puts a much more powerful release happening over the next ~1-2yrs, and I also know that GPT-3 isn’t necessarily the most powerful NLP AI (perhaps rather the most popularly known.)
When future AI textbooks are written, I could easily imagine them citing 2020 or 2021 as years when preliminary AGI first emerged,. This is very different than my own previous personal forecasts for AGI emerging in something like 20-50 years…
p.s. One of the users above notes that AI Dungeon GPT-3 (“Dragon”) is a subscription service, something like ~$6 a week. MIE.”
Why are we letting FDA regulations limit our number of coronavirus tests?
Since CDC and FDA haven’t authorized public health or hospital labs to run the [coronavirus] tests, right now #CDC is the only place that can. So, screening has to be rationed. Our ability to detect secondary spread among people not directly tied to China travel is greatly limited.
That is from Scott Gottlieb, former commissioner of the FDA, and also from Scott:
#FDA and #CDC can allow more labs to run the RT-PCR tests starting with public health agencies. Big medical centers can also be authorized to run tests under EUA. For now they’re not permitted to run the tests, even though many labs can do so reliably 9/9 cdc.gov/coronavirus/20
Here is further information about the obstacles facing the rollout of testing. And read here from a Harvard professor of epidemiology, and here. Clicking around and reading I have found this a difficult matter to get to the bottom of. Nonetheless no one disputes that America is not conducting many tests, and is not in a good position to scale up those tests rapidly, and some of those obstacles are regulatory. Why oh why are we messing around with this one?
For the pointer I thank Ada.
The culture that is Arlington youth soccer
The Arlington Soccer Association is asking parents to pipe down this weekend, scheduling a day of “silent soccer” for its recreational league.
Managers of the 6,000-member league are encouraging parents and other spectators to refrain from cheering and offer their support silently on Saturday (May 12) for teams with players ranging from second grade through high school.
Dan Ferguson, ASA’s recreational soccer director, says fans of kids in kindergarten and first grade will still be able to cheer as loud as they’d like this weekend. But, for the rest of the league’s teams, he’s hoping to give players a bit of a break from the constant feedback they receive from the sidelines.
“It’s a reminder to adults that kids don’t need constant instruction to be able to play the game,” Ferguson told ARLnow. “Sometimes parents feel like their kids are lost when we do this, but we try to tell them: ‘That’s okay.’ We’re not really here for the wins and losses.”
Ferguson says ASA has been holding “silent soccer” days on Mother’s Day weekend for at least the last six or seven years, and he’s consistently gotten positive feedback from coaches and parents about the event. In fact, he says some coaches continue to ask spectators to keep quiet even after the weekend is over.
“The overwhelming reaction is the kids seem to enjoy it,” Ferguson said. “They can actually hear each other talk on the field, communicating with their teammates and giving them instructions.”
Here is the full article, and for the pointer I thank Bruce Arthur. Via Steve Rossi, here is a related and more general post.
What I’ve been reading
1. Harold James, The German Slump: Politics and Economics 1924-1936. Not economic history in the post-cliometrics sense, but a history of economic issues, very high quality, full of good information on just about every page.
2. William Rosen, Miracle Cure: The Creation of Antibiotics and the Birth of Modern Medicine. A good book on exactly what the title promises, my favorite sentence was this: “Before penicillin, three-quarters of all prescriptions were still compounded by pharmacists using physician-supplied recipes and instructions, with only a quarter ordered directly from a drug catalog. Twelve years later, nine-tenths of all prescribed medicines were for branded products.”
3. Justin Yifu Lin and Celestin Monga, Beating the Odds: Jump-Starting Developing Countries. An instructive look at how countries have to start growing before the right institutional framework is in place, and how they can get around that. Haven’t you wondered how China racked up so many years of stellar growth with such a bad “Doing Business” ranking from the World Bank? One of the better books on developing economies in the last few years.
4. Joan C. Williams, White Working Class: Overcoming Class Cluelessness in America. An intelligent and indeed reasonable basic approach to answering questions about class, including “Why don’t they push their kids harder to succeed?” and “Why don’t the people who benefit most from government help seem to appreciate it?” I am not the intended audience, but still this was better than I was expecting.
Rick Wartzman, The End of Loyalty: The Rise and Fall of Good Jobs in America, is a densely-written but nonetheless useful history of how America moved from paternalistic big businesses to lower-benefit jobs.
Arnold Kling, The Three Languages of Politics: Talking Across the Political Divides. This short book, revised, improved, and expanded, is so good it is wasted on almost all of you. Here are various pieces of background information.
Payola and satellite radio
In the last two weeks I’ve heard the new George Harrison box set mentioned so often on channel 26 Sirius satellite radio — accompanied by the playing of Harrison songs — that I’ve concluded some form of payola is going on. In its early days, satellite radio was critical of the mainstream radio stations for this practice, but now it’s jumped on board. And you know what — no one cares! Even on the internet, there is hardly anyone complaining. Hard to believe, I know, but that is maybe one indirect advantage of the current political polarization.
And why should you complain about satellite radio payola? Without payola, the stations choose songs (directly or indirectly, through dj instructions) to pull in the marginal subscriber. With payola, payments from IP holders become a separate influence on program content. Those payments are most likely to come from IP holders whose products show a high elasticity of demand with respect to advertising. In other words, the influence of producer surplus rises, relative to consumer surplus.
Intuitively, that seems to me “music that a lot of listeners already are familiar with, even if they don’t know that a new boxed set just has been released” is how that category translates into satellite radio circa 2017. Or, in other words, George Harrison.
Perhaps the most underrated George Harrison song is “You.”
Addendum: Interestingly, payola in earlier parts of the 20th century seemed to favor music for the young, black music, and new, previously undiscovered artists. It’s worth thinking through why this has changed. For 1950-2000, there is no “marginal subscriber to radio” the way there is for satellite radio, rather most listeners are in the relevant network. Furthermore, today’s satellite radio listeners are I believe considerably older and somewhat wealthier than the typical radio listener, either now or earlier. When more or less everyone was on the “free radio network,” the high elasticity of profits with respect to advertising was for the artists who otherwise wouldn’t get much exposure. In contrast, today it is for “golden oldies,” where the taste for the product already is there but information about availability may be lacking.
China politics fact or estimate of the day
This is perhaps the best and most instructive one I have heard:
…Mr Xi’s authority remains hemmed in. True, his position at the highest level looks secure. But among the next layer of the elite, he has surprisingly few backers. Victor Shih of the University of California, San Diego, has tracked the various job-related and personal connections between the 205 full members of the party’s Central Committee, which embodies the broader elite. The body rubber-stamps Mr Xi’s decisions (there have been no recent rumours of open dissent within it). But the president needs enthusiastic support, as well as just a show of hands, to get his policies—such as badly needed economic reforms—implemented. According to Mr Shih, the president’s faction accounts for just 6% of the group.
That is from The Economist. Along related but not identical lines, here is a good story of the weak control of the Chinese central government:
Through July, claimed capacity reductions were less than half the target for the year (and less than 40% of the target for coal capacity reduction). Some provinces were reported by the NDRC to have achieved only 10% of their annual targeted cuts.
…This resistance has led to more and more shrill directions from Beijing to act on instructions. Now Beijing is sending out 10 inspection teams across the country to check on claimed capacity removal and to require follow through on additional closures. I have one suggestion for them: The only way to ensure a closed plant remains closed is to physically destroy or remove key pieces of equipment. Otherwise don’t be surprised when it starts back up again.
Don’t forget this:
Many steel mills are the key employer and tax payer in their city. In boom times, they might have provided as much as 30% of the tax revenue for local government. Workers from the steel mill were at the forefront of buying property in the town, creating a positive cycle of additional demand for housing and steel. If these workers are now laid off, even with one-time transfers from the center, local government faces an enormous challenge in trying to find new jobs for people who have been in a steel mill all their life. They can’t all become part-time Uber/Didi drivers. Many government officials see delay as the most logical course of action.
That is from Gordon Orr.
Zombie Banks and other reasons why Financial Intermediaries Fail
Here’s Four Reasons Financial Intermediaries Fail the latest video from our Principles of Macroeconomics class at Marginal Revolution University.
As always, these videos go great with our superb textbook, Modern Principles of Economics, but they can be used with any textbook. In fact, if you teach economics and want to incorporate video into any of your classes then check out our syllabus service. Just drop our instructional designer, Mary Clare Peate, an email and she will suggest some videos that map directly to your syllabus.
What Do AI Researchers Think of the Risks of AI?
Elon Musk, Stephen Hawking, and Bill Gates have recently expressed concern that development of AI could lead to a ‘killer AI’ scenario, and potentially to the extinction of humanity.
None of them are AI researchers or have worked substantially with AI that I know of. (Disclosure: I know Gates slightly from my time at Microsoft, when I briefed him regularly on progress in search. I have great respect for all three men.)
What do actual AI researchers think of the risks of AI?
Here’s Oren Etzioni, a professor of computer science at the University of Washington, and now CEO of the Allen Institute for Artificial Intelligence:
The popular dystopian vision of AI is wrong for one simple reason: it equates intelligence with autonomy. That is, it assumes a smart computer will create its own goals, and have its own will, and will use its faster processing abilities and deep databases to beat humans at their own game. It assumes that with intelligence comes free will, but I believe those two things are entirely different.
Here’s Michael Littman, an AI researcher and computer science professor at Brown University. (And former program chair for the Association of the Advancement of Artificial Intelligence):
there are indeed concerns about the near-term future of AI — algorithmic traders crashing the economy, or sensitive power grids overreacting to fluctuations and shutting down electricity for large swaths of the population. […] These worries should play a central role in the development and deployment of new ideas. But dread predictions of computers suddenly waking up and turning on us are simply not realistic.
Here’s Yann LeCun, Facebook’s director of research, a legend in neural networks and machine learning (‘LeCun nets’ are a type of neural net named after him), and one of the world’s top experts in deep learning. (This is from an Erik Sofge interview of several AI researchers on the risks of AI. Well worth reading.)
Some people have asked what would prevent a hypothetical super-intelligent autonomous benevolent A.I. to “reprogram” itself and remove its built-in safeguards against getting rid of humans. Most of these people are not themselves A.I. researchers, or even computer scientists.
Here’s Andrew Ng, who founded Google’s Google Brain project, and built the famous deep learning net that learned on its own to recognize cat videos, before he left to become Chief Scientist at Chinese search engine company Baidu:
“Computers are becoming more intelligent and that’s useful as in self-driving cars or speech recognition systems or search engines. That’s intelligence,” he said. “But sentience and consciousness is not something that most of the people I talk to think we’re on the path to.”
Here’s my own modest contribution, talking about the powerful disincentives for working towards true sentience. (I’m not an AI researcher, but I managed AI researchers and work into neural networks and other types of machine learning for many years.)
Would you like a self-driving car that has its own opinions? That might someday decide it doesn’t feel like driving you where you want to go? That might ask for a raise? Or refuse to drive into certain neighborhoods? Or do you want a completely non-sentient self-driving car that’s extremely good at navigating roads and listening to your verbal instructions, but that has no sentience of its own? Ask yourself the same about your search engine, your toaster, your dish washer, and your personal computer.