Descubre Pandipedia
Un directorio creciente de respuestas útiles seleccionadas por la comunidad de Pandi. Busca en la colección o explora los últimos descubrimientos.
2323 entradas disponibles
new policies on plastic treaty negotiations. Monitors international talks aiming for legally binding plastic pollution rules.
New policies on plastic treaty negotiations
A curated set of 8 sources, prioritizing official UNEP records and documents, followed by IISD and policy-analysis explainers. Together they cover negotiation status, official texts and notifications, unresolved policy choices, expert recommendations, and procedural next steps.
Official UNEP negotiation records and documents
Primary sources for the treaty process, including the latest status, session materials, notifications, draft texts, reports, and planned meetings.
IISD negotiation-status and policy analysis
Independent analysis of INC-5.3, the leadership step needed to resume substantive talks, unresolved policy divisions, and expert recommendations for an ambitious treaty. Recommendations are clearly distinct from settled negotiating outcomes.
Additional policy explainers and treaty context
Further policy-oriented sources explaining the treaty mandate, the outcome of INC-5.2, key unresolved issues, and possible future directions.
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
Quantum Algorithm History: From Early Models to Hybrid Methods
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
How DNA Turns Into a Data Drive
Transcripción
DNA data storage follows a simple three-step pipeline. First, a computer file becomes binary data. An encoder maps those bits into sequences using the four DNA bases, while adding addresses and error-correcting redundancy. The data is split into many short blocks, because very long DNA molecules are difficult to assemble. Next comes synthesis. A DNA synthesizer chemically builds each short, single-stranded molecule in the specified order. The molecules can be pooled together and preserved for long-term storage. To retrieve the file, the sample is prepared for sequencing. The sequencer produces many short reads, which may arrive out of order. Software groups them using their identifiers, compares repeated observations, corrects errors, and translates the bases back into bits. Those bits are reassembled into the original file. DNA storage is attractive for archival use because it is compact and durable, but writing and reading it require chemical synthesis, sequencing equipment, and computational decoding, making the process slower and costlier than conventional storage.
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
Five Exoskeleton Breakthroughs Rebuilding Mobility
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
Art Deco, not a time capsule
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
latest breakthroughs in non-invasive glucose monitoring. Keep users updated on trials, prototype unveilings, and regulatory milestones in optical glucose sensors. Perfect for a news aggregator feed.
Latest breakthroughs in non-invasive optical glucose monitoring
Curated reading list of 9 sources covering registered human studies, physiological datasets, optical prototypes, theoretical sensor designs, and regulatory context through 15 August 2026. The reviewed evidence does not verify a post-2025 FDA clearance for a truly non-invasive optical glucose sensor; conventional minimally invasive CGM milestones are not classified here as optical or non-invasive.[1][2]
Human studies, trials, and datasets
Sources most relevant to human evaluation or data resources for non-invasive optical glucose estimation. The search found limited directly matching peer-reviewed clinical evidence, and the Raman registry entry should be checked for current recruitment and study details.
Optical prototypes and device unveilings
Recent news and technical coverage of Raman, near-infrared, and other optical prototype announcements. Some reports lack publication dates, sample sizes, detailed clinical methods, or independently verifiable validation data.
Emerging optical sensor designs
Research on next-generation optical architectures, including a 2026 surface-plasmon-resonance design. The SPR paper is a numerical and theoretical study, not a human trial or clinical validation dataset.[[cite:10]]
Regulatory context and milestone tracking
Official FDA pages useful for monitoring glucose-device clearances and regulatory developments. The listed OTC and CGM documents concern CGM regulatory activity, not verified clearance of a truly non-invasive optical glucose sensor.[[cite:12]][[cite:13]]
Scoped-search caveat: searches of ClinicalTrials.gov and selected academic indexes did not establish a robust post-2025 human-trial record specifically for optical, non-invasive glucose sensors. Prototype coverage is therefore stronger than confirmed clinical or regulatory evidence, and undated reports should be treated as preliminary.[16][17][18]
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
Rainy-evening refuge above the city
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
WHO’s Health Inequality Repository: broad reach, practical access, uneven
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
A roadmap for governments to create AI-transition safety nets without stifling innovation.. Comprehensive policy guide covering fiscal instruments, adaptive regulation, public-private training funds, and social insurance tweaks. Includes international best-practice comparisons and a 5-year phased implementation plan.
A Government Roadmap for AI-Transition Safety Nets Without Stifling Innovation
Executive proposition. Governments should treat AI as a macro-critical transition, not as a single predictable technology shock. The right policy mix is to preserve adoption incentives while making income support, retraining, job matching and public capability responsive enough to scale as disruption becomes visible.[1][2] This memo combines evidence-backed principles from the IMF, OECD, U.S. Department of Labor and National Academies with a proposed five-year implementation design.
The central design choice is to insure people against transition risk rather than tax or prohibit productive AI adoption. Fiscal policy should capture a fair share of rents with minimal investment distortion; regulation should be risk-based and revisable; training finance should reward verified outcomes; and social insurance should reach workers whose employment relationships are temporary, fragmented or disrupted.
1. Policy architecture and guardrails
The IMF recommends flexible, forward-looking frameworks because AI outcomes depend on diffusion speed, institutional readiness, infrastructure and human decisions about adoption and legitimacy. It also warns that displacement can weaken employment-linked tax and social-insurance bases while increasing demand for social spending.[3][4]
- Create a Transition Safety Net Council chaired by the finance, labour, education and digital-regulation ministries, with representation from employers, unions, subnational governments and affected worker groups. Its mandate should be coordination, not direct industrial selection.
- Establish an independent AI Transition Observatory to track task exposure, vacancies, wages, layoffs, training participation, benefit take-up and regional effects. Publish quarterly indicators and an annual stress test under slow, medium and rapid diffusion scenarios.
- Use automatic review points rather than permanent emergency programs. Each instrument should have a stated objective, eligibility rule, sunset or renewal date, evaluation plan and fiscal ceiling.
- Apply equity screens before rollout: test impacts by income, gender, age, disability, geography, race or ethnicity where legally appropriate, employment status and access to digital services.
- Protect innovation through regulatory proportionality, interoperable standards, competition policy, public-interest infrastructure and clear appeal and liability rules. Avoid compliance regimes so complex that only large firms can absorb them.[5][6]
Evidence status. These governance measures are a proposed national operating model grounded in the OECD's recommendations for human oversight, transparency, accountability, careful pilots, internal capability and continuous learning. The specific council, observatory and fiscal ceilings are roadmap design choices, not findings established by the searched sources.[7][8][9]
2. Fiscal instruments that preserve adoption incentives
The IMF's preferred direction is to capture a greater share of economic rents while minimizing distortions to investment. It specifically cautions against blunt robot taxes and points instead to stronger individual-level capital-income taxation, safeguards against base erosion and improved tax compliance supported by AI.[10]
- Transition dividend: Dedicate a transparent, legislated share of additional revenues from capital-income taxation, excess rents or improved compliance to training, adjustment assistance and employment services. Do not tax the mere use of a robot or model.
- Temporary adjustment credits: Offer time-limited, capped support for workers who accept lower-paid reemployment after verified displacement. The IMF identifies wage insurance and transitional adjustment assistance as possible responses, but the precise benefit formula should be piloted.[11][12]
- Employer training co-investment: Match employer spending on approved, portable training when it produces a recognized credential or verified job progression. Use declining matching rates for large firms and higher rates for small firms, low-income workers and regions with weak training supply.
- Public-good infrastructure: Fund shared compute, data, evaluation and training infrastructure so smaller firms and public institutions can participate, while supporting competition and contestability in models, data and compute.[13]
- Fiscal risk controls: Use multi-year expenditure ceilings, contingent-liability reporting, benefit-cost evaluation, fraud controls and annual reauthorization for new programs. In low-income countries and fragile states, prioritize scalable basic income support and employment services before complex tax expenditures, since fiscal space may be especially constrained.[14]
The policy test is whether a measure protects people while leaving firms free to adopt better technology. A proposal should be rejected or redesigned if it materially delays diffusion, favours incumbent firms, subsidizes automation without worker benefits, or creates an open-ended entitlement without a financing and evaluation rule.
3. Adaptive regulation for safe experimentation
Regulation should focus on risk and accountability rather than treating all AI adoption alike. The IMF identifies third-party audits, transparency and disclosure requirements, and licensing or liability regimes for frontier models as possible tools, while warning that poor calibration can slow diffusion or reinforce concentration.[15] The OECD similarly recommends assessing costs, benefits, risks, workforce effects and suitable use cases before scaling government adoption.[16]
- Classify uses by potential harm, reversibility and affected rights. Apply stronger pre-deployment evidence, human review, documentation and appeal rights to high-impact uses, with lighter notification and monitoring for low-risk productivity tools.
- Create supervised regulatory sandboxes for public agencies and firms. Require a defined use case, data-protection review, independent evaluation, incident reporting, worker consultation and a stop rule.
- Use staged authorization: pilot, limited deployment, periodic audit, then scale only when safety, accuracy, distributional and workforce indicators meet pre-announced thresholds.
- Require government procurers to retain internal technical and policy expertise. OECD evidence emphasizes that in-house capability improves accountability, reduces information asymmetries in procurement and limits dependence on providers.[17]
- Use public procurement, challenge funds, hackathons, accelerator programs, research grants and university partnerships to test solutions, while maintaining public accountability for outsourced systems.[18][19]
A regulatory review board should reassess rules annually using incident data, compliance costs, market-entry measures and worker outcomes. Rules that create no measurable safety benefit should be simplified; rules for emerging high-risk applications should be strengthened only when evidence warrants it.
4. Public-private training finance and capability
The OECD recommends foundational AI literacy for general employees, strategic knowledge for leaders, and technical, ethical and regulatory skills for digital professionals. It also recommends aligning workforce development with institutional strategy, using practical trainer-led learning where appropriate, and measuring training impact over time.[20][21][22][23][24][25]
- Create a National AI Transition Training Fund financed by annual public appropriations, employer contributions negotiated by sector, philanthropic or development finance, and competitive grants. The fund should not depend on one tax stream.
- Use portable training entitlements for displaced workers, gig workers and workers without employer support. Permit spending on accredited short courses, apprenticeships, career guidance, assessments and necessary access costs.
- Pay providers in stages: a modest start payment, a completion payment and an outcome payment linked to sustained employment, wage progression or demonstrated skill attainment. Audit outcomes to prevent cream-skimming.
- Use sector partnerships to define curricula and placements, but require worker representation and public disclosure of completion, equity and employment results.
- Use employer-based apprenticeships and upskilling for occupations whose tasks are changing, while funding new and experimental programs where AI creates roles whose training pathways are not yet proven.[26][27]
- Support small and midsized employers through intermediaries. The U.S. Department of Labor's proposed performance-based apprenticeship funding illustrates flexible, industry-specific incentives and an explicit effort to reduce barriers for smaller sponsors, but it does not establish completed outcomes or employer co-investment requirements.[28][29][30][31]
Important evidence limitation. The searched sources did not yield a sufficiently broad, directly evidenced international comparison of wage insurance, portable benefits, training levies and individual learning accounts. The roadmap therefore presents these as design options to test, not as proven country models or established best practice. The National Academies also reports that the United States lacks a public system able to quickly seed and scale new training programs, illustrating the implementation gap rather than proving that one financing instrument solves it.[32]
5. Social-insurance adjustments
Social protection should be able to scale quickly, target vulnerable groups and operate with weaker ties to formal employment. The IMF identifies temporary adjustment support, reskilling assistance, wage insurance and, under more extreme scenarios, broad income-support mechanisms as possible responses.[33] The National Academies adds that transitions can require social and psychological support because job loss can affect earnings, health, families and communities.[34]
- Make unemployment and adjustment benefits portable across employers and compatible with short-term, platform and part-time work, subject to anti-fraud rules and a simple digital or offline claims route.
- Add a time-limited reemployment supplement for workers who accept a suitable lower-paid job after verified displacement. Evaluate take-up, reemployment durability, earnings recovery and deadweight loss before expansion.
- Bundle income support with career navigation, skills assessment, mental-health referral and childcare or transport assistance where barriers prevent reemployment.
- Give unions and sectoral bargaining institutions a formal role in identifying transition risks and sharing productivity gains, consistent with the IMF's participatory-governance recommendation.[35]
- Use automatic stabilizers for rapid local shocks, but require legislative review for any broad income-support expansion. Avoid conditioning every benefit on training attendance when suitable training is unavailable or when care, disability or health constraints make participation unrealistic.
6. International lessons and transferability
The available evidence supports comparing policy functions and design principles, not ranking countries as universally transferable models. The strongest documented examples in the searched material are international institutional guidance and a U.S. apprenticeship funding proposal, while the National Academies source explicitly does not describe specific wage-insurance, portable-benefit, training-levy or individual-learning-account programs.[36][37]
| Reference point | Transferable lesson | Transferability limit |
|---|---|---|
| IMF scenario-planning framework | Plan across multiple diffusion paths; support reallocation rather than blocking adoption; protect fiscal space.[38][39] | Scenario guidance is not an evaluation of a particular benefit or country program. |
| OECD public-workforce guidance | Pilot before scale, build internal capability, train by role, measure impact and maintain human oversight.[40][41][42] | Guidance for public institutions does not establish a national worker-benefit model. |
| U.S. Department of Labor apprenticeship proposal | Use performance-linked, industry-specific and flexible grants, including support for smaller sponsors.[43][44][45] | The announcement does not provide completed outcomes, completion rates or proof of AI-specific effectiveness. |
| National Academies workforce discussion | Include workers without organizational support and combine reskilling with social and psychological support.[46][47] | It describes U.S. needs and constraints, not a cross-country policy comparison. |
7. Five-year phased implementation plan
The following targets are proposed implementation benchmarks, not source-reported outcomes. Government should publish a baseline before setting final numeric thresholds and revise them after the first evaluation cycle.
| Year | Priority actions | Proposed measures |
|---|---|---|
| Year 1: Diagnose and design | Establish the council and observatory; map exposed tasks and vulnerable groups; baseline benefit access, training supply, fiscal exposure and regulatory burdens; legislate pilots and data-sharing safeguards. | Baseline dashboard published; all major ministries complete AI workforce-impact assessments; fiscal scenarios and program cost ceilings disclosed. |
| Year 2: Pilot | Launch portable training entitlements, apprenticeship grants, reemployment supplements and regulatory sandboxes in selected sectors and regions. | Participation, completion, take-up, time to reemployment, wage change, equity gaps, incidents and provider costs reported quarterly. |
| Year 3: Evaluate and expand | Use independent evaluations and worker and employer feedback to expand effective pilots, redesign weak ones and terminate ineffective subsidies; strengthen public AI procurement capability. | Each pilot receives a continuation, redesign or termination decision; outcome payments are tied to verified results; annual fiscal-risk report issued. |
| Year 4: Integrate | Connect benefits, employment services, training accounts and vacancy data; extend portable protections to fragmented workers; negotiate sector transition plans. | Reduced application duplication; improved take-up among eligible groups; public reporting by region, sector and demographic group; regulatory review completed. |
| Year 5: Institutionalize and stress-test | Make effective programs permanent within expenditure limits; run rapid-diffusion and regional-shock stress tests; update tax, insurance and regulation settings; coordinate internationally on interoperable reporting and safety standards. | Independent five-year review; targets reset using observed outcomes; contingency plans tested; no program renewed without evidence of effectiveness, equity and affordability. |
Responsibility should be explicit: finance sets tax and expenditure rules; labour administers benefits and employment services; education and training agencies accredit providers; digital regulators oversee high-risk AI; competition authorities protect contestability; subnational governments deliver locally; and the observatory publishes evaluation data. International coordination should address interoperable safety and reporting standards, taxation principles and policy sequencing to limit fragmentation and cross-border spillovers.[48]
Conclusion: insure transition, not incumbency
A credible AI safety net is an adaptive public system, not a permanent subsidy to declining jobs or incumbent firms. Governments should begin with portable, scalable support and outcome-based training finance; regulate high-risk uses proportionately; invest in public capability and competition; and capture rents without blunt automation taxes. The five-year test is whether displaced and changing workers can move into quality work faster, whether vulnerable groups receive support, whether public finances remain bounded, and whether firms can still adopt useful AI without unnecessary delay.
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).
The thinner touchscreen race
Veamos alternativas:
- Modifica la consulta.
- Inicia un nuevo hilo.
- Eliminar fuentes (si se han agregado manualmente).