Chat With Turk, Outline

How ChatWithTurk used to work:

  1. A user says a phrase A to Turk
  2. Turk remembers phrase A for later.
  3. Turk thinks of some ‘similar’ things that Turk has previously said that resemble A. (These are potential phrases B)
  4. For each of B, Turk checks to see what responses he has received when Turk said those. Turk picks one of these. This is phrase C. (If Turk has no good historic B phrases, he uses an untried one, something he’s heard but never said, which is his phrase C.)
  5. Turk responds with phrase C, which hopefully shares some context with phrase A, or maybe is a wild guess.

Right off the bat we have a system that has a lot of inherent randomness, even though it doesn’t have any entropy – the page just collects user input and regurgitates it according to the above. It does get tuned with use, insofar as the list of phrases (and appropriate human given responses) grows over time. Of course, Turk doesn’t track context at all, and doesn’t even differentiate from the conversation to conversation on his own. From close up, it’s very naive.

However, it works as sort of a conversational echo chamber – the user dictates the course of the conversation, whether they are greeting the Turk (he often returns the greeting), insulting him (he usually responds in kind to profanity), or asking questions about the nature of the page. He often accuses the user of being a bad chatbot.

At best, the conversations generated by a mature Turk system more closely resemble the ones found in the front cover of an old yearbook than a live conversation, and that makes sense. Since doing this experiment, I’ve let it go, but email me for free source code.

You should chat with a Turk

There’s a little Internetology experiment I’d like to share with you. What if I paired up a random person on the Internet for a conversation? Actually, that’s been done before.

What if I paired you up with a random person, but told you that it was a chatbot? Finally, what if that chatbot was non-stateful and just replied the best it could to everything you said to it? Would that the result of that be interesting or just confusing?

(Coarse language warning: Like I said, this is an experiment in chatting with strangers on the Internet…)

Try Chat with A Turk. Go ahead, I’ll post the results and an overview of the algorithm here soon.

Things the Collective Should Do: Open Academic Web Journal

I’m suggesting here a free-content, openly accessible online repository where researchers, professors, and students publish scientific journal articles for peer review and wide distribution.

This website allows scientific research papers to be published by qualified academicians. The articles can be freely read and peer-reviewed online. Articles are translated into other languages so that they can be read worldwide. A system of of moderation provides a meritocratic means of awarding prestige and press based on quality. An accompanying print journal is provided pro bono or at nominal cost to institutions. To increase the prestige of the journal/repository, it is marketed as a trustable, progressive, intelligent institution, and content is carefully reviewed.

The current scientific publication industry relies on established branding of respectable journals and the ‘publish or perish’ dynamic to keep it afloat. Authors often have to pay for their articles to be published in print-bound journals, which are then sold at a high price to academic institutions. At best the publishing industry contributes little of value to the system, and at worst prevents most people from accessing information that could be useful in the hands of the general public. In essence, the current system lacks utility in spreading scientific knowledge and neither apportions prestige fairly, nor distributes knowledge widely.

Minds around the world would benefit greatly, as the results of studies would be available internationally in many different languages. Universities and authors would also benefit because they are now able to publish and access papers at lower cost. Science as a whole would benefit due to the increased volume and visibility of papers published. Because of the greater number of eyes on the articles and the increased ease of peer review, communicative openness and the scientific method would benefit.

In order to be successful, the open web journal would require buy-in from academic institutions and scientific readers. A combination of aggressive marketing and branding to entice article submissions will facilitate presenting the site as a respectable, reliable source of information. We will need to develop the website’s software, decide how the site is run and edited organizationally (peer review and editing will play a huge part)

The overall progress of science will be assisted, because knowledge will be exchanged more freely. People who would otherwise not have the opportunity to read current scientific literature will have the chance to be inspired as well as educated by it. Competing with current journal models may persuade existing publishers to become more free in an economic and cultural sense. Researchers will have a website which will both distribute their knowledge to the world and grant them recognition for their work – without charging admission. Counting readers or articles would be simple metrics. Measuring changes in research job satisfaction, number of articles published worldwide, or cost of subscriptions to existing print journals would tell other sides of the story.

Things the Collective Should Do: Developing World B2B Portal

Let’s imagine a website which gives small businesses in developing areas around the world access to free web-based business software, financial planning resources, commercial contacts, and other information they need to be successful.

This website will provide leaders of new and developing businesses access to a suite of software that provides simple and reliable double entry bookkeeping, payroll processing, sales forecasting, and other necessary business management functions. Additionally, the website will include a free portal to crucial services such as employment postings, regional legal guidelines, sales opportunities, and business development tools. Access to local and global contacts via this portal will give businesses more viability, give them access to wider markets, and put them in touch with the contacts they need to be successful. Businesses may be able to apply for microfinanced loans, government incentives, and other means of attaining capital.

This website will address informational and logistical barriers to entry of small businesses in remote or underdeveloped areas without an existing support network, widespread infrastructure, and preexisting culture of entrepreneurship. The

Entrepreneurs in the developing world would have better access to verified information, support networks, and new markets. This empowerment of individual businesses would also lead to new local jobs, increased availability of goods and services, and the generation of wealth throughout communities.

The first step to properly implementing this website would be to conduct a needs assessment of areas that would benefit from online business development software to figure out who we could reach and what we could offer them. Creating usable, reliable online business management software with the features needed by small businesses in the developing world, translated for many possible markets would follow, and finally, recruiting business experts to share knowledge of standard practices, financial howtos, and entrepreneurial techniques would get the site up and functioning.

Optimally, those businesses that used the website would have better access to crucial information and would be more likely to succeed and grow. These businesses would set an example in their communities and share their knowledge via the website, enabling sustainable business practices to spread through their localities. One would compare the outcomes of a sample of businesses who used the software with similarly equipped businesses in the region who did not. For product-driven companies, an increase in the number of potential customers who would like to import their product would lead to fairer, more competitive pricing for goods.

Things the Collective Should Do: Free Online College

We’re talking completely free, accredited, online associate-level college courses.

A free online college offering courses adhering to high levels of quality. The courses are (in US terms) in the 100-300 level range, enough to satisfy general education requirements and demonstrate a student’s commitment to an institution of higher education. Courses are accredited by relevant standards bodies, and admission is not restricted in any way. Students’ knowledge of the course topic is vetted by open-book online tests, peer reviewed papers, peer-judged class competitions, readings followed by captcha-like comprehension tests, class discussion followed by peer ratings, and/or other scalable systems of measuring comprehension of materials. Courses are copyleft and the subject matter is crowdsourced and peer reviewed.

Cyclical educational disparity exists worldwide. Entirely-online classes are becoming increasingly common but still have costs that preclude the enrollment of the average world citizen . Education is not a zero-sum game, and information is easily recyclable for many minds. Everyone should have a chance at achieving a high level of education, and this idea removes some of the social and economic barriers to this.

Marginalized populations who currently do not have access to high school or college would be able to obtain a higher education, and would be able to apply to other colleges with proof of their academic experience. Existing colleges would have a much wider pool of applicants from more diverse backgrounds. Additionally, seniors and working adults will have the opportunity to engage in life-long learning. Society worldwide will be enriched by a general increase in education.

The first step is the creation of software that would allow people to freely contribute to open sourced course material. The success of Wikipedia is indicative of the willingness of Internet users to contribute information and editing to worthwhile causes. All course content would be reviewed by Professors to ensure that accreditation standards are being met. The software in question would also be usable by students. The college would require marketing designed to appeal to a diverse student body. Peer editing and adherence to high levels of quality will facilitate wide spread accreditation of classes.

If more people obtain free college-level learning, the idea will be a success. The disparity within education can be measured in terms of average levels of learning across overlapping boundaries of gender, income, nationality, and race. The number of person-credit-hours would impart the degree of success of the free online college. A shift in international attitudes toward learning would also indicate improvement. A measurement of the educational divide will be demonstrated in an increased proportion of college students from marginalized backgrounds.